100 Ans Enseignement Mathematiques

One Hundred Years of LEnseignement Math matique e
Moments of Mathematics Education in the Twentieth Century
Proceedings of the EMICMI Symposium Geneva, 2022 October 2000
edited by Daniel CORAY, Fulvia FURINGHETTI, H l` ne GISPERT, Bernard R. HODGSON and Gert SCHUBRING ee
LENSEIGNEMENT MATHEMATIQUE
Gen` ve 2003 e
Tous droits de traduction et de reproduction r serv s pour tous pays e e ISSN : 0425-0818 ISBN : 2-940264-06-6
c 2003 LEnseignement Math matique, Gen` ve e e
TABLE OF CONTENTS
List of participants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Authors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . INTRODUCTION : DANIEL CORAY BERNARD R. HODGSON ......
5 6 9
LENSEIGNEMENT MATHEMATIQUE : BIRTH AND STAKES

FULVIA FURINGHETTI Mathematical instruction in an international perspective : the contribution of the journal LEnseignement Math matique . . . . . . . . e GERT SCHUBRING LEnseignement Math matique and the rst International Commise sion (IMUK) : the emergence of international communication and cooperation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . GILA HANNA Journals of mathematics education, 19002000 . . . . . . . . . . . . . REACTION : JEAN-PIERRE BOURGUIGNON GENERAL
DISCUSSION
19
47 67 85 89
...................
(reported by CHRIS WEEKS) . . . . . . . . . . . .
GEOMETRY
RUDOLPH BKOUCHE La g om trie dans les premi` res ann es de la revue LEnseignement e e e e Math matique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e GEOFFREY HOWSON Geometry : 195070 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . COLETTE LABORDE G om trie P riode 2000 et apr` s . . . . . . . . . . . . . . . . . . . . . . e e e e REACTION : NICOLAS ROUCHE GENERAL
DISCUSSION
95 113 133 155 160
...........................
(reported by MARTA MENGHINI) . . . . . . . . .
EMICMI SYMPOSIUM
ANALYSIS
JEAN-PIERRE KAHANE Lenseignement du calcul diff rentiel et int gral au d but du e e e vingti` me si` cle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e e MAN-KEUNG SIU Learning and teaching of analysis in the mid twentieth century : a semi-personal observation . . . . . . . . . . . . . . . . . . . . . . . . . . . LYNN STEEN Analysis 2000 : challenges and opportunities . . . . . . . . . . . . . . . ` REACTION : MICHELE ARTIGUE GENERAL
DISCUSSION
167
179 191 211 224
..........................
(reported by MARTA MENGHINI) . . . . . . . . .
APPLICATIONS OF MATHEMATICS : MATHEMATICS AS A SERVICE SUBJECT

PHILIPPE NABONNAND Les d bats autour des applications des math matiques dans les e e r formes de lenseignement secondaire au d but du vingti` me si` cle e e e e ` HELENE GISPERT Applications : les math matiques comme discipline de service dans e les ann es 19501960 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . e MOGENS NISS Applications of mathematics 2000 . . . . . . . . . . . . . . . . . . . . . REACTION : GERHARD WANNER . . . . . . . . . . . . . . . . . . . . . . . . . . GENERAL
DISCUSSION
229
251 271 285 293
(reported by CHRIS WEEKS) . . . . . . . . . . . .
PERSPECTIVES FOR MATHEMATICS EDUCATION

UBIRATAN DAMBROSIO Stakes in mathematics education for the societies of today and tomorrow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . JEREMY KILPATRICK Scientic solidarity today and tomorrow . . . . . . . . . . . . . . . . . . REACTION : HYMAN BASS ..............................
301 317 331
LIST OF PARTICIPANTS
N stor AGUILERA (Argentina) e Mich` le ARTIGUE (France) e Nataliya BANTSUR (Ukraine) Hyman BASS (USA) Rudolph BKOUCHE (France) Jean-Pierre BOURGUIGNON (France) Aldo BRIGAGLIA (Italy) Daniel CORAY (Switzerland) Ubiratan DAMBROSIO (Brazil) Viviane DURAND-GUERRIER (France) Trevor FLETCHER (UK) Ruhal FLORIS (Switzerland) Fulvia FURINGHETTI (Italy) Claude GAULIN (Canada) Livia GIACARDI (Italy) H l` ne GISPERT (France) ee Alexandra GOMES (Portugal) Francine GRANDSARD (Belgium) Pierre-Paul GRIVEL (Switzerland) Andr HAEFLIGER (Switzerland) e Ernst HAIRER (Switzerland) Gila HANNA (Canada) Jean-Claude HAUSMANN (Switzerland) Bernard R. HODGSON (Canada) Geoffrey HOWSON (UK) Jean-Pierre IMHOF (Switzerland)
Djordje KADIJEVIC (Yugoslavia) Jean-Pierre KAHANE (France) Jeremy KILPATRICK (USA) Colette LABORDE (France) Thomas LINGEFJARD (Sweden) Marta MENGHINI (Italy) Alain MERCIER (France) Philippe NABONNAND (France) Mogens NISS (Denmark) Jo o PITOMBEIRA DE CARVALHO (Brazil) a Lus REIS (Portugal) Andr REVUZ (France) e Nicolas ROUCHE (Belgium) Lee RUDOLPH (USA) Jean RUFFET (Switzerland) Gert SCHUBRING (Germany) Mansa C. SINGH (Canada) Man-Keung SIU (China) Per SIVERTSEN (Norway) Lynn STEEN (USA) John STEINIG (Switzerland) Alain VALETTE (Switzerland) Gerhard WANNER (Switzerland) Claude WEBER (Switzerland) Chris WEEKS (UK) Jochen ZIEGENBALG (Germany)
THE AUTHORS
Mich` le ARTIGUE : e IREM, Case 7018, Universit Paris 7 Denis Diderot, 2, place Jussieu, F-75251 e Paris Cedex 05, France [artigue@math.jussieu.fr] Hyman BASS : 4204C School of Education, 610 E. University, University of Michigan, Ann Arbor, MI 48109-1259, U. S. A. [hybass@umich.edu] Rudolf BKOUCHE : 64, rue N grier, F-59800 Lille, France [rbkouche@wanadoo.fr] e Jean-Pierre BOURGUIGNON : Institut des Hautes Etudes Scientiques, 35, route de Chartres, F-91440 Bures-surYvette, France [jpb@ihes.fr] Daniel CORAY : Section de math matiques, Universit de Gen` ve, 24, rue du Li` vre, Case postale e e e e 240, CH-1211 Gen` ve 24, Switzerland [Daniel.Coray@math.unige.ch] e Ubiratan DAMBROSIO : R. Peixoto Gomide, 1772 Apt 83, 01409-002 Sao Paulo SP, Brazil [ubi@usp.br] Fulvia FURINGHETTI : ` Dipartimento di Matematica, Universita di Genova, Via Dodecaneso 35, I-16146 Genova, Italy [furinghe@dima.unige.it] H l` ne GISPERT : ee Groupe dHistoire et de Diffusion des Sciences dOrsay (GHDSO), B atiment 407, Centre universitaire, F-91405 Orsay Cedex, France [Helene.Gispert@ghdso.u-psud.fr] Gila HANNA : Ontario Institute for Studies in Education, University of Toronto, 252 Bloor St. West, Toronto, Ontario M5S 1V6, Canada [ghanna@oise.utoronto.ca] Bernard R. HODGSON : D partement de math matiques et de statistique, Universit Laval, Qu bec e e e e G1K 7P4, Canada [bhodgson@mat.ulaval.ca] Albert Geoffrey HOWSON : 1 Chidden Holt, Chandlers Ford, GB-Eastleigh SO53 4RJ, United Kingdom [geoffreyhowson@aol.com] Jean-Pierre KAHANE : 11, rue du Val-de-Gr ce, F-75005 Paris, France a [Jean-Pierre.Kahane@math.u-psud.fr]
EMICMI Symposium
Jeremy KILPATRICK : 105 Aderhold Hall, University of Georgia, Athens GA 30602-7124, U. S. A. [jkilpat@coe.uga.edu] Colette LABORDE : Laboratoire IMAG-Leibniz, Universit Joseph Fourier CNRS, 46, avenue F lix e e Viallet, F-38000 Grenoble, France [Colette.Laborde@imag.fr] Marta MENGHINI : ` Dipartimento di Matematica, Universita di Roma La Sapienza, Piazzale Aldo Moro, I-00185 Roma, Italy [menghini@mat.uniroma1.it] Philippe NABONNAND : Archives Henri Poincar , Universit de Nancy 2, 23, Boulevard Albert 1er, BP 3397, e e F-54015 Nancy Cedex, France [nabonnan@isis.univ-nancy2.fr] Mogens NISS : IMFUFA, Roskilde University, P.O. Box 260, DK-4000 Roskilde, Denmark [mn@mmf.ruc.dk] Nicolas ROUCHE : CREM, 5, rue Emile Vandervelde, B-1400 Nivelles, Belgium [rouche@math.ucl.ac.be] Gert SCHUBRING : Institut f r Didaktik der Mathematik, Universitat Bielefeld, Postfach 100 131, u D-33501 Bielefeld, Germany [gert.schubring@uni-bielefeld.de] Man-Keung SIU : Department of Mathematics, University of Hong Kong, Hong Kong SAR, China [mathsiu@hkucc.hku.hk] Lynn Arthur STEEN : Department of Mathematics, St. Olaf College, Northeld MN 55057, U. S. A. [steen@stolaf.edu] Gerhard WANNER : Section de math matiques, Universit de Gen` ve, 24, rue du Li` vre, Case postale e e e e 240, CH-1211 Gen` ve 24, Switzerland [Gerhard.Wanner@math.unige.ch] e Chris WEEKS : Downeycroft, Virginstow Beaworthy, Devon EX21 5EA, United Kingdom [chrisweeks@eurobell.co.uk]
Charles-Ange LAISANT (18411920)
INTRODUCTION
Organized jointly by the University of Geneva and the International Commission on Mathematical Instruction, a symposium on mathematics education was held in Geneva in October 2000 under the theme
ONE HUNDRED YEARS OF LENSEIGNEMENT MATHEMATIQUE : TH MOMENTS OF MATHEMATICS EDUCATION IN THE 20 CENTURY
It was an occasion for celebrating the 100th anniversary of the international journal LEnseignement Math matique, founded in 1899 by Henri Fehr e (Geneva) and Charles-Ange Laisant (Paris). Among periodicals devoted to mathematics education, the new journal was the rst to seek explicitly an international audience ; in fact, an original characteristic of the beginnings was a series of articles on the teaching of mathematics in different countries. The fact that such a journal was launched in Geneva is not surprising if we consider the local context in which it emerged. Apart from obvious references to internationalism as seen from Henry Dunants birth-place, it is also true that, for centuries, Geneva had been haunted by the myth of its pedagogic vocation. To describe this strong historical background, it is useful to recall here that the city of Geneva was not a part of the Swiss Confederation until 1815. It had been an episcopal principality for several hundred years with a relatively large degree of autonomy granted to the people until it became an independent protestant republic, by a decree of 21st May 1536, in which the Conseil g n ral promulgated the Reformation 1 ). It is quite remarkable e e that the same decree set the principle of compulsory education, which was also declared free for the poor :
` [] que chescung soit tenu envoyer ses enfans a lescholle et les faire apprendre. 2 ) [Encyclop die de Gen` ve 1986, 127] e e
1 ) To avoid a common confusion, it is important to note that this decision was reached under the determining inuence of Guillaume Farel, a few months before Jean Calvin arrived in Geneva. 2
) [] that everyone be required to send their children to school and have them learn.
10
INTRODUCTION
In 1559 Jean Calvin inaugurated two new institutions : the Coll` ge (with e seven grades) and the Acad mie. Led by Mathurin Cordier, Th odore de B` ze, e e e and others, both schools attracted numerous students from France, the German States, the Low Countries and the whole of Europe, right from the beginning 3 ) but even more after the Saint-Barth lemy massacre (1572). e Mathematics and physics were also taught. However, a formal chair of mathematics was introduced in the Acad mie only in 1724, with Jean-Louis e Calandrini and Gabriel Cramer as professors. In spite of these auspicious beginnings, many questions were raised throughout the 18th century about the quality, the content and the methods of instruction. The best known name in this connection is that of Jean-Jacques Rousseau, who was concerned with pedagogy 4 ). But the discussion, which involved famous scientists like Horace-B n dict de Saussure, went on more e e specically about the respective roles of the sciences and the humanities 5 ), or the creation of technical schools. At the beginning of the 19th century, many educators were active on this fertile soil. One was Rodolphe T pffer, a professor of Greek and rhetoric in o Genevas Acad mie who created his own pedagogic system and even opened e a boarding school. He is also viewed by some as the inventor of the comic strip. The cartoons selected here 6 ) illustrate, with a specic hint at mathematics, the kind of animated debate that was going on about the organization of primary and secondary schools (public or private) at that time. They are drawn from Monsieur Cr pin [T pffer 1837, 1112], a comic strip published e o in Geneva in 1837. They depict a father who is trying to nd a suitable preceptor for his children. In the preceding cartoon one of the rst teachers he has hired explains that all pupils now begin to proceed very well from the general to the particular. An example is provided by the cartoon about Besancon. The father is of course rather unhappy, even though the answer is mathematically correct. The next cartoon shows how the system can fail if the pupil is not particularly gifted. Considering the early date (1837), any allusion to Bourbaki and the new maths movement in the third cartoon would be sheer anachronism. But the questions were already very clearly stated.
3 4
) John Knox, Jean de L ry, and other distinguished names, were among them. e ) Emile, ou de l ducation was published in the Hague and Paris in 1762, and burnt in e
Geneva in that same year !

5 ) Saussure published a Projet de r forme pour le Coll` ge in 1774; cf. [Encyclop die de e e e Gen` ve 1988, 1314]. e 6 ) reproduced by the kind permission of Editions du Seuil.
INTRODUCTION
11
The teacher summons Joseph and asks him : Where is Besancon ? Joseph instantly replies that Besancon is in the set of all things, which includes the Universe, which includes the World, which includes all four parts of the World, which include Europe, where Besancon lies.
Having called Leopold, Mr. Cr pin himself asks how e much will eight pounds of lard cost, at ve orins a pound. Leopold instantly replies that lard is in the set of all things, which includes the Universe, which includes all three reigns, which include the animal reign, which includes the pig, which includes lard.
Mr. Cr pin nds that his son e Leopold is little advanced in Arithmetic. The teacher explains that, in his system, Arithmetic is the very last thing that Leopold will know. Indeed, he must rst know Algebra, which he will begin to learn only after an in-depth study of quantity in general.
The Acad mie was itself undergoing a process of transformation which e was giving it a status more in line with that of other universities in Europe 7 ). Since Fehr began his studies and completed his doctorate in Geneva, he was certainly aware of these discussions. In connection with Furinghettis article in this monograph, we can also mention that, as a professor in the University, he had Th odore Flournoy and Edouard Clapar` de as colleagues. e e The rst prefaces of LEnseignement Math matique show that Fehr and e Laisant also wanted to associate the world of teaching to the great movement of scientic solidarity which was emerging at the end of the 19th century,
7 ) The Acad mie evolved into the University of Geneva in 1872, with a Faculty of Medicine, e independence from the clergy, etc.
12
INTRODUCTION
notably through the organization of international meetings such as the rst International Congress of Mathematicians held in Zurich in 1897. The journal immediately obtained important successes, as is testied by the gold medal at the World Fair of Brussels in 1905. The idea of internationalism in mathematics education was crucial to the journal right from its beginning and it even led to some articles in or about Esperanto. Moreover the frequent advocacy in the journal of the importance of an international perspective played an essential role in the establishment, a few years later, of the International Commission on Mathematical Instruction. The two editors Fehr and Laisant had proposed in 1905 to organize an international survey on the reforms needed in mathematics education, asking in particular for opinions on the following theme : les conditions que doit remplir un enseignement complet, th orique et pratique, des math matiques e e dans les etablissements sup rieurs 8 ) [La R daction 1905, 382]. e e In response to a question on the progress needed in the organization of the teaching of pure mathematics, the US mathematician and teacher educator, David Eugene Smith, explicitly suggested the establishment of an international commission to study the situation :
Pour ce qui est de la premi` re question, jestime que la meilleure mani` re de e e renforcer lorganisation de lenseignement des math matiques pures, serait de e cr er une commission qui serait nomm e par un Congr` s international et qui e e e etudierait le probl` me dans son ensemble. 9 ) [Smith 1905, 469] e
Smith repeated this suggestion in a report he presented at the Fourth International Congress of Mathematicans, held in Rome in April 1908. This Congress then adopted a resolution 10 ) to the effect of appointing a committee, composed of Felix Klein (Germany) as President, George Greenhill (Great Britain) as Vice-President, and Henri Fehr (Switzerland) as Secretary-General, with the mandate to constitute an International Commission to organize a comparative Study on the methods and plans of mathematics teaching in secondary schools. This International Commission eventually developed a much wider scope of interest and became the ICMI as we know it today. The coincidence of the spirit of internationalism between the newly established Commission on teaching and the journal LEnseignement Math e
8 ) The conditions to be satised by a complete theoretical and practical teaching of mathematics in higher institutions. 9 ) As regards the rst question, I consider that the best way to reinforce the organization of the teaching of pure mathematics, would be the establishement of a committee appointed by an international Congress and which would study the problem in its entirety. 10
) See in this connection LEnseign. Math. 21 (1920), 306.
INTRODUCTION
13
matique, possibly combined with the presence of Fehr being active in both groups, resulted in one of the very rst decisions of ICMI being to select the journal as its ofcial organ. This was announced simply as follows in the journal as part of a report on the decision of the Rome Congress to establish an International Commission : LEnseignement math matique servira dorgane e ` ` a la Commission, dont la t che se rattache tr` s intimement a celle que poursuit a e notre revue internationale depuis dix ans. 11 ) [Chronique 1908, 333] The explicit mention of the journal as the ofcial organ of ICMI can be seen on the front cover of the rst issue following the inception of the Commission (volume 11, 1909). And it has been appearing on the cover ever since then, except for some periods when ICMI was inactive around the Second World War. Another distinguishing feature of LEnseignement Math matique at the e outset was a keen interest of the journal in the social role of mathematics and of science in general. Laisant in particular was the author of several articles on these topics. To give an idea of his fairly optimistic vision, we may quote : Si la science pouvait devenir exclusivement utilitaire, elle perdrait sa plus ` grande utilit . 12 ) [Laisant 1904, 342], and also : Mesurer une science a son e utilit est presque un crime intellectuel. 13 ) [Laisant 1907, 121]. e The aim of the symposium organized by the University of Geneva and ICMI was to look at the evolution of mathematics education over the twentieth century and to identify some guidelines and trends for the future, taking into account, among other sources, the documents, debates and related papers having appeared in LEnseignement Math matique. The emphasis was on e secondary education (students in the age range of about 12 to 18 or 19 years) and also included the education of teachers. The programme of the symposium was based on a series of invited talks. The Programme Committee 14 ) had identied three main themes to be discussed geometry, analysis, and applications of mathematics and three different periods at which these themes were to be considered : 1900, 1950 (i.e. the period leading to the new maths), and 2000.
11 ) LEnseignement Math matique will serve as organ of the Commission, whose task is very e intimately linked to the aim our international journal has been pursuing for ten years. 12 13 14
) If science could become exclusively utilitarian, it would lose its greatest usefulness. ) Measuring a branch of science by its usefulness is something like an intellectual crime. ) The members of the Programme Committee of the Symposium were Daniel Coray
(Switzerland), Fulvia Furinghetti (Italy), H l` ne Gispert (France), Bernard R. Hodgson (Canada) ee and Gert Schubring (Germany).
14
INTRODUCTION
It was also a gathering of some of the main actors, during the last decades, in mathematics education as considered from an international perspective, and ample time was devoted during the symposium to collective discussions on the themes presented in the talks. In this connection some participants had been invited to play the role of reactors. They had the responsibility, in each session, of launching the discussion following the invited talks, partly by giving a synthesis of the presentations but more importantly by outlining the major trends and issues about the theme, both in the light of the past century and as seen from todays perspective. This explains the subdivision adopted in this book. We must add that every contribution, in either English or French, is preceded by a fairly long abstract in the other language. We hope these Proceedings will contribute to a better understanding and appreciation, among the communities of mathematicians and mathematics educators, of the evolution of mathematics education during the 20 th century. The book that we are proposing to the reader aims at reecting the spirit and the work of a symposium which, in the words of Geoffrey Howson, demonstrated
how over the century the emphasis shifted from discussions of the mathematics to be taught to an elite, to the needs of a wider range of students and of society. [] It reminded us of the way in which two generations had tried to make enormous changes in the content of school mathematics and methods of teaching it. It gave us an opportunity to see where these earlier efforts had not been wholly successful and challenged us to determine why. With such an understanding we should be better equipped to tackle both the problems that now face us and those which will arise in the future. [Howson 2001, 183]
This endeavour would not have been possible without the generous support of the following institutions : Swiss Academy of Sciences, Swiss National Science Foundation, Commission Administrative de lUniversit de e Gen` ve, Facult des Sciences (University of Geneva), and Soci t Acad mique e e ee e (Geneva). We are very grateful to all of them. We are also much indebted to Chris Weeks and to Marta Menghini for their spontaneous help, particularly for transcribing and editing reports of the discussions. Daniel Coray University of Geneva (Switzerland) President of the Foundation LEnseignement Math matique e Bernard R. Hodgson Universit Laval (Qu bec, Canada) e e Secretary-General of ICMI
INTRODUCTION
15
REFERENCES CHRONIQUE. Commission internationale de lenseignement math matique. LEnseign. e Math. 10 (1908), 333. ` ENCYCLOPEDIE DE GENEVE. Tome 5 (Les religions). Association de lEncyclop die de e Gen` ve, Gen` ve, 1986. e e Tome 6 (La science et l cole). Association de lEncyclop die de Gen` ve, Gen` ve, e e e e 1988. HOWSON, G. A. A report on the International Symposium organised jointly by the University of Geneva and ICMI. LEnseign. Math. (2) 47 (2001), 181183. LAISANT, C.-A. Le r le social de la science. LEnseign. Math. 6 (1904), 337362. o La Math matique, philosophie, enseignement. (2e ed.) Gauthier-Villars, Paris, e 1907. ` LA REDACTION. Note de la R daction sur les r formes a accomplir dans lenseignement e e des math matiques. LEnseign. Math. 7 (1905), 382383. e ` SMITH, D. E. Opinion de David-Eugene Smith sur les r formes a accomplir dans e lenseignement des math matiques. LEnseign. Math. 7 (1905), 469471. e TOPFFER, R. Monsieur Cr pin. Gen` ve, 1837. Reprinted in : R. T pffer, Monsieur e e o Cr pin, Monsieur Pencil : Deux egarements de la science. Editions du Seuil, e Paris, 1996.
Henri FEHR (18701954)
MATHEMATICAL INSTRUCTION IN AN INTERNATIONAL PERSPECTIVE : THE CONTRIBUTION OF THE JOURNAL LENSEIGNEMENT MATHEMATIQUE L ducation math matique dans une perspective internationale : e e la contribution de la revue LEnseignement Math matique e par Fulvia FURINGHETTI
e La revue LEnseignement Math matique a et fond e en 1899, dans un moment de e e grand ferment social et culturel qui touchait notamment les math matiques. Dune part e les Etats s taient dot s dune organisation moderne comportant, entre autres, la mise e e en place de syst` mes nationaux dinstruction. Dautre part, la recherche math matique e e s tait d velopp e dans diff rentes directions (pures ou appliqu es), et ce de mani` re e e e e e e tr` s efcace. Plusieurs revues de recherche math matique etaient alors r guli` rement e e e e publi es dans divers pays et le premier Congr` s international des math maticiens avait e e e e et organis a Zurich en 1897. Dans le domaine de lenseignement des math matiques, e` e des associations nationales denseignants avaient vu le jour dans de nombreux pays et e e des revues avaient et cr ees portant uniquement sur lenseignement de cette discipline. Les id es centrales faisant de LEnseignement Math matique un journal remarquable e e dans ce panorama sont linternationalisation et la communication. Les fondateurs de la revue le Suisse Henri Fehr (18701954) et le Francais Charles-Ange Laisant (18411920) en furent les directeurs jusqu` leur mort. Dans la pr sentation du a e journal publi e dans le premier num ro, ils ecrivaient que le monde de lenseignement e e devait sassocier au grand mouvement de solidarit scientique . Ils soulignaient e aussi limportance des r formes des programmes et du probl` me de la formation des e e enseignants. Pour aborder ces questions, ils pr conisaient que lon compare les syst` mes e e dinstruction dans les diff rents pays et que les enseignants echangent leurs points de e vue. Le pr sent texte porte sur les premi` res ann es du journal, jusqu` la premi` re e e e a e Guerre mondiale, au cours desquelles le projet des fondateurs a trouv un terrain e
20
F. FURINGHETTI
` favorable a son d veloppement. A cette epoque, le Comit de patronage du journal e e comprenait des personnages importants du milieu math matique (directeurs de revues e math matiques, historiens, math maticiens), tous avec un int r t marqu pour les e e ee e probl` mes li s a l ducation math matique. Le journal publiait e e ` e e des articles g n raux, e e des nouvelles du monde acad mique, e des annonces bibliographiques et des comptes rendus de livres, articles et conf rences, e des correspondances diverses, des enqu tes lanc es par le journal, ainsi que les r sultats accompagn s de e e e e commentaires. La langue utilis e alors est quasi exclusivement le francais. Les articles g n raux e e e abordent des th` mes math matiques dint r t pour lenseignement aux niveaux secone e ee daire et tertiaire, ainsi que lhistoire, la philosophie, l pist mologie et la psychologie. e e Les auteurs des articles et des lettres etaient pour la plupart francais ou suisses, mais on trouvait aussi des contributions provenant dautres pays (Alg rie, Allemagne, Are gentine, Autriche et les r gions de son ancien empire, Belgique, Danemark, Espagne, e Etats-Unis, Gr` ce, Italie, Japon, Pays-Bas, Portugal, Roumanie, Royaume-Uni, Russie, e Ukraine). D j` au cours de la premi` re ann e et durant les ann es subs quentes, on ea e e e e trouve dans la revue des articles sur la situation de lenseignement des math matiques e dans divers pays, en conformit avec les buts dinformation et de communication du e journal. ` La correspondance permit l tablissement de contacts entre les lecteurs et fut a e lorigine dinitiatives int ressantes. Par exemple, cest une lettre dun lecteur qui inspira e une enqu te lanc e par le journal sur les m thodes de travail des math maticiens. Cette e e e e enqu te fut r alis e au moyen dun questionnaire de 30 questions adress es aux lecteurs e e e e et autres math maticiens int ress s. Le questionnaire fut pr par par Fehr avec laide e e e e e de deux psychologues de lUniversit de Gen` ve, Edouard Clapar` de et Th odore e e e e Flournoy. Le journal publia le questionnaire ainsi que lanalyse des r ponses et des e commentaires sur les r sultats, r dig s par Fehr, Clapar` de et Flournoy. Cette enqu te e e e e e ` contribua a attirer lattention des math maticiens (entre autres, Henri Poincar ) et des e e enseignants sur certains th` mes reli s a la psychologie et sur le probl` me de linvention e e ` e en math matiques. e ` La discussion sur les r formes a accomplir dans lenseignement des math matiques, e e lanc e en 1905 par des lettres de math maticiens (parmi lesquels David E. Smith et e e ` ` Gino Loria), fut a lorigine dun mouvement plus g n ral qui aboutit a la fondation e e de la Commission internationale de lenseignement math matique (CIEM/ICMI) lors e ` du Congr` s international des math maticiens tenu a Rome en 1908. A partir de 1909, e e le journal devint dailleurs lorgane ofciel de la CIEM et on y publia r guli` rement e e les annonces, actes des rencontres et enqu tes de la Commission. Fehr fut Secr taire e e g n ral de la CIEM depuis sa cr ation jusquau d but des ann es 50. e e e e e
MATHEMATICAL INSTRUCTION IN AN INTERNATIONAL PERSPECTIVE : THE CONTRIBUTION OF THE JOURNAL LENSEIGNEMENT MATHEMATIQUE by Fulvia FURINGHETTI
INTRODUCTION Among all the mathematical journals of the nineteenth century LEnseignement Math matique plays a particular role 1 ). It was founded in 1899 e by Charles-Ange Laisant and Henri Fehr and is still published today 2 ). To understand why its role is so interesting I will analyse the rst years of its life, in which it made a strong contribution to the birth of an international community of mathematics educators, alongside the community of mathematics researchers. We shall see that this analysis provides us with the opportunity to rethink important aspects of the history of mathematical instruction, including the birth of ICMI.
1 ) Appendix 1 gives an idea of the range of the mathematical press when LEnseignement Math matique was founded. See [Friedelmeyer 1996] for further information on early mathematical e journals. 2 ) At rst the publishers were G. Carr and C. Naud in Paris (since 1902 C. Naud alone). e In 1904 the administration and the printing house were transferred to Geneva, and the publishers were Gauthier-Villars (Paris) and Georg & Co. (Geneva). The present publisher is an independent foundation linked to the Department of Mathematics of the University of Geneva. The printer has been KUNDIG (Geneva) since 1903, with slight variations of the commercial name : W. K undig & ls, Albert K ndig, Albert Kundig (umlaut removed in 1914 !), etc., now SRO-Kundig. u
22
F. FURINGHETTI
In the rst issue of LEnseignement Math matique the editors explained e the aims of the journal in the following terms :
A lheure o` la science a tant progress , certaines simplications peuvent u e etre d sirables, les programmes des diverses branches de lenseignement e appellent des r formes plus ou moins compl` tes. Et avec cela, il y a une e e question fondamentale dont on ne saurait m connatre limportance : cest celle e de la pr paration du corps enseignant. e Toutes ces transformations ne sauraient saccomplir brusquement, ni sans ` de s rieuses r exions pr alables. Mais, pour proc der a une telle etude dune e e e e facon judicieuse et utile, la premi` re des conditions nest-elle pas de connatre ce e qui se passe dans les autres pays, de savoir quel est dans chacun deux le mode dorganisation de lenseignement math matique, quels sont les programmes en e vigueur, les moyens de sanction des etudes, etc. ? [] Malgr les relations e ` fr quentes qui se sont etablies a notre epoque entre savants qui cultivent e un m me sujet d tude, malgr les congr` s internationaux, si brillamment e e e e inaugur s a Zurich en 1897 et dont le principe est d sormais etabli, le e ` e monde de lenseignement proprement dit na pu sassocier jusqu` pr sent a e ` a ce grand mouvement de solidarit scientique aussi pleinement quil e t et e u e d sirable. [] e Nous avons voulu, par la publication de notre Revue, renverser les obstacles ` qui sopposent a ces communications r ciproques et cr er une sorte de e e correspondance mutuelle, continue, entre les hommes qui ont consacr leur e ` vie a cette noble mission : l ducation math matique de la jeunesse. e e e ` En vue de ce r sultat, notre premier soin a et de donner a la publication e p riodique dont il sagit un caract` re franchement et hautement international. 3 ) e e [Les Directeurs 1899, 12]
The target readership of the journal was those who taught mathematics (at secondary and tertiary level). As stated by the directors in the introduction to the sixth volume of LEnseignement Math matique : e
3 ) At a time when science has made so much progress, some simplications may be desirable, the programmes of the different branches of teaching require more or less complete reforms. Linked to this there is a fundamental issue whose importance should not be overlooked : namely that of the education of teachers. All these changes ought not to be carried out in an abrupt way, nor without serious preliminary reections. But is not the rst requirement for proceeding in such a study in a judicious and fruitful way, to be aware of what happens in other countries, to know the way mathematics teaching is organised in each of them, what teaching programmes are in force, the methods by which the studies are approved, etc. ? [] Despite the strong rapport that has been established today among scientists of the same eld, despite the international congresses of mathematicians, so brilliantly inaugurated in Zurich in 1897 and accepted as a principle for the future, the world of education proper has not up to now been able to join this great movement of scientic solidarity as fully as would have been desirable []. It has been our wish, through the publication of our Journal, to overcome the obstacles to reciprocal communications [my italics] and to create a kind of continuous mutual correspondence between men who have devoted their lives to this noble mission : the mathematical education of young people. In view of this aim, our rst concern has been to give this periodical a clearly and openly international [my italics] character.
MATHEMATICAL INSTRUCTION IN AN INTERNATIONAL PERSPECTIVE
23
Le mot Enseignement a pour nous la signication la plus large. Il veut dire e enseignement des el` ves, et aussi enseignement des professeurs et dailleurs lun ne va gu` re sans lautre. 4 ) [LEnseign. Math. 6 (1904), 4] e
To consider those who teach as persons accomplishing a noble mission, rather than persons following a profession, has been a typical attitude of the past. The social role of science and links with progress in its various forms (industry, technology, ) are among the concerns of the editors. The key words of the journals programme are internationalism, information, communication and, of course, teaching mathematics. Professional mathematicians were invited to collaborate actively with the project of the journal in order to keep the teaching of mathematics in touch with the advances in the subject itself. Behind the ideas of internationalism and solidarity expressed in the journal there lay social and political ideals, very much alive in society at the time of its foundation, but which slowly withered away in the new century. The following passage hints at the changes in the international atmosphere and their consequences for the life of the journal :
La guerre europ enne porte un coup sensible aux institutions internationales. e Dans les pays bellig rants et dans les pays neutres voisins tout ce que la nation e compte dhommes valides est sous les drapeaux. Il devient donc mat riellement e ` impossible de continuer les travaux faisant appel a de nombreux collaborateurs. ` Les uvres de paix telles que la n tre passent a larri` re plan. Dailleurs, o e poursuivant un id al commun librement choisi, elles exigent une volont dunion e e quon ne saurait demander aux savants dans une p riode aussi troubl e que e e celle que nous traversons. 5 ) [Fehr 1914, 477]
I consider 1914 (the year of the beginning of the First World War) as the limit of the journals pioneering period, in which it found a propitious environment in the international social atmosphere. My analysis of the journal goes 6 ) from the year of the foundation (1899) to this crucial year 1914.
4 ) The word Enseignement [teaching] has for us the widest possible meaning. It means the teaching of pupils, as well as the teaching of teachers and, indeed, you can hardly have the one without the other. 5 ) The European war carries with it an appreciable impact on international institutions. In the ghting countries and in the neighbouring neutral countries all those whom the nation considers to be able-bodied are in uniform. Thus it becomes practically impossible to continue with activities requiring the support of numerous collaborators. Peaceful activities like ours have to take second place. Furthermore, being in pursuit of a freely chosen common ideal, they demand a willingness for unity, which ought not to be asked of scientists in a period as troubled as the one in which we are living now. 6
) The steps of the history of the journal until the 1970s are sketched in [de Rham 1976].
24
F. FURINGHETTI
THE
FOUNDERS
As is well known, early mathematical journals were strongly linked to their founders, to the extent that many of them were referred to as the journal of its founder. For example the Journal fur die reine und angewandte Mathematik (founded by August Leopold Crelle in Berlin, rst issue in 1826) was simply known as Crelles Journal. This was also the case for the Journal de math e matiques pures et appliqu es (founded by Joseph Liouville in Paris, rst issue e ` in 1836) and the Giornale di Matematiche ad uso degli studenti delle universit a italiane (founded by Giuseppe Battaglini in Naples, rst issue 1863). In some cases the editors were also the owners of the journal. In journals devoted to mathematics teaching the importance of the editorial line is twofold : on the one hand cultural (in connection with mathematics), on the other hand social (in connection with systems of instruction). For the period we are considering, the editorial line of LEnseignement Math matique may be seen e as a real expression of the personalities of its two founders and editors. Charles-Ange Laisant was born in Basse-Indre (France) on 1st November 1841 and died in Paris (1920). In his obituary he is described as homme de science, educateur, philosophe et politicien [Buhl 1920, 73] (see also [Sauvage 1994 ; Pascal 1983]). Both his life and his contributions to the mathematical community are a combination of all these characteristics. He studied at the Ecole polytechnique in Paris. His rst mathematical works were on the application of the theory of equipollence and the method of quaternions (see [Ortiz 1999 ; 2001]). Afterwards he turned predominantly to social and instructional themes. His view of society, based on values such as solidarity, collaboration and communication stimulated his enterprises in the community of mathematicians. In 1894 he founded (with Emile Lemoine) the journal LInterm diaire des Math maticiens, which was conceived as a e e means of providing contact between mathematicians through the exchange of questions (which were something more than mere exercises) and answers, together with bibliographical references. In the rst issue of that journal [1 (1894), question 212, 113] the idea of the organisation of international congresses of mathematicians, to be held at regular intervals, was explicitly launched. Laisant also edited the Annuaire des Math maticiens (C. Naud, e Paris, 1902), a publication that contained the names and addresses of living mathematicians, of scientic societies and of scientic periodicals. He was, with Poincar , a member of the commission charged with the production of e the R pertoire bibliographique des sciences math matiques (a precursor of the e e present journals of mathematical reviews).
25
In the eld of mathematical education and instruction Laisants works on pedagogy of mathematics were greatly valued by his contemporaries in France and abroad. His educational and philosophical leanings are evident in the editorial line of the journal and in the contributions he made to LEnseignement Math matique. For example, Buhl [1920] writes that Laisants ideas e were imbued with the philosophy of Auguste Comte and, indeed, in the second year of LEnseignement Math matique, we nd an article on the philosophy of e mathematics of Comte (see [Vassilief 1900]). Interesting aspects of Laisants personality are shown by his contributions to French political life. Ortiz [2001] reports that he was a member of the national parliament in the Third Republic, up to the end of the century. In France this was a period of social changes which had repercussions in the scientic world. It was held that advancement in science occurs according to criteria similar to those necessary for industrial development. As Ortiz put it, Laisant showed
cierta coherencia entre las agendas cientca y poltica [], ambas muestran una preocupaci n seria por introducir ideas nuevas, por abrir nuevas posibilidades, o y por buscar unidad dentro de una aparente diversidad. 7 ) [Ortiz 2001, 83]
Henri Fehr was born in Zurich on 2nd February 1870 and died in Geneva (1954). He studied in Switzerland and afterwards in France. His doctoral dissertation was on the method of Grassmann vectors applied to differential geometry. Jacques Hadamard gave a positive review of this work. Fehr became professor at the University of Geneva (Sciences Faculty) and later dean, vicerector and rector. Further information on Fehr is in [Anonymous 1955 ; de Rham 1955 ; Ruffet 1955]. A prominent characteristic of Fehrs personality was his interest in the social aspects of the mathematical community and academic life. He was regarded as a p dagogue exceptionnel [Anonymous 1955, 7], but he was e also involved in social commitments such as the committee of the fund of pensions of his colleagues. He applied his skill as an organiser in founding 8 ) the Swiss Mathematical Society (of which he was president), the Foundation for the Advancement of Mathematical Sciences, and the journal Commentarii Mathematici Helvetici. He received national and international honours and appointments. Fehr participated, as his countrys delegate, in the International Congresses of Mathematicians and was vice-president of the ICMs in Toronto
7 ) clear consistency between his political and scientic behaviours. Both in politics and in sciences he was seriously concerned with the introduction of new ideas, of opening up new possibilities, and looking for unity in things that appear different. 8
) with R. Fueter and M. Grossmann
26
F. FURINGHETTI
(1924) and Bologna (1928). He was also vice-president of IMU. He was one of the founders of ICMI (in 1908), of which he was the secretary-general until the Second World War. When ICMI was reconstituted (in Rome, 1952) Fehr was one of the members of the special committee of ve appointed to draw up a plan of work in preparation for the International Congress of Mathematicians to be held in Amsterdam (in 1954). When this committee coopted new members to form an ICMI Executive Committee, Fehr was again chosen to be the honorary president of this new committee (see [Behnke 19511954]). Throughout all these years Fehr was the real soul of ICMI. The main contribution of Fehr as author of articles in the journal was in the eld of mathematics instruction. He singled out some central points, such as :
innovations in the mathematical programmes and their links with the
development of science and technology, as an echo of what was happening in many countries (see [Schubring 1996]) ;
the relationship between pure and applied mathematics and its inuence
on the mathematics teaching ;

the education of mathematics teachers ; new trends in mathematics teaching.
From a reading of what has been written about Laisant and Fehr one is left with the impression that these two men enjoyed the esteem of all those who had contact with them. As a pair they are a good example of the integration of interests and strengths, which was essential to the enterprise of editing LEnseignement Math matique. e
THE
FORMAT OF THE JOURNAL
LEnseignement Math matique was published every two months. Initially e (see [Les Directeurs 1899]) all papers were published in French and this continued to be the predominant language, despite the fact that the editorial address of 1913 states that papers written in the ofcial languages of the International Congresses of Mathematicians (English, French, German and Italian) were accepted, with Esperanto also allowed. One of the very few articles appearing in a non-French language (followed by a long summary in French) is The principles of mathematics in relation to elementary teaching, the text of the talk delivered at the International Congress of Mathematicians in Cambridge (1912) by A. N. Whitehead [1913]. The admission of Esperanto
27
was in line with the atmosphere of internationalism advocated by the editors and the strong interest in international languages shown by the mathematical milieu (see [Roero 1999]). Actually a few contributions to the journal were written in international languages, two of them by Charles M ray [1900 ; 1901] e illustrating the advantages of Esperanto for the internationalism of science, one by M. Frechet [1913]. From 1903, the French mathematician Adolphe Buhl (18781949) was added as collaborator to the editorial board and, after Laisant died, Buhl became one of the editors of the journal and held that ofce until his death 9 ). Until 1914 the journal had a Comit de patronage. The international nature of the journal e can be gauged from the names of the members of the rst Comit (1899) : e
Paul Appell, Paris Nicolas Bougaiev (Bougajeff), Moscow (until 1903) Moritz Benedikt Cantor, Heidelberg Luigi Cremona, Rome (until 1903) Emanuel Czuber, Vienna Zoel Garca de Galdeano y Yanguas, Zaragoza Alfred George Greenhill, Woolwich, England Felix Klein, G ttingen o Valerian Nikolajwitsch Liguine (Ligin), Warsaw (until 1900) Paul Mansion, Gent G sta Magnus Mittag-Lefer, Stockholm o Gabriel Oltramare, Geneva (until 1906) Julius Peter Christian Petersen, Copenhagen (until 1910) Emile Charles Picard, Paris Henri Jules Poincar , Paris (until 1912) e Pieter Hendrik Schoute, Groningen (until 1913) Kyparissos Stephanos, Athens Francisco Gomes Teixeira, Porto Alexandr Wassiljewitsch Vassilief (Wassilief), Kazan Alexander Ziwet, Ann Arbor, Michigan
The following new members took the place of deceased members : in 1904 Vasiliy Petrovich Ermakof (Ermakoff), Kiev ; Andrew Russell Forsyth, Cambridge (until 1910) ; Gino Loria, Genoa ; David Eugene Smith, New York ; in 1907 J r me Franel, Zurich 10 ). The composition of the Comit covered a eo e wide range of abilities and interests including research in mathematics or in the history of mathematics, editing journals, the writing of books, teacher education and mathematics instruction. From 1905 the journal carried the subtitle M thodologie et organisation e de lenseignement. Philosophie et histoire des math matiques. Chronique e scientique M langes Bibliographie. Following the inauguration of ICMI e (in Rome, 1908) the lives of ICMI and of the journal were intertwined ; since 1909 the frontispiece of the journal has described itself as the Organe ofciel
9 10
) For further information on Buhl see [Fehr 19421950]. ) The Comit de patronage ceased to exist in 1915. e
28
F. FURINGHETTI
de la Commission internationale de lEnseignement math matique 11 ). e The normal practice (both in the past and today) is for journals to contain a set of articles and a small part devoted to selected announcements and reviews of books and articles appearing elsewhere. In LEnseignement Math matique e the main articles were only a part of the journal, the other sections being just as important. In 1904, when the administration and the printing services moved from Paris to Geneva, the editors clearly explained the structures of the journal and the purposes of the different sections. They stressed again their intention to foster communication between researchers and teachers. As an example of the format of the journal, we can look at the organisation of the material published in 1905 (the seventh year of the journal, the middle of the period under consideration). We nd the following sections :
A) Articles g n raux (General articles : Methodology and organisation of teaching, e e History and Philosophy) B) M langes (Miscellanies) e C) Correspondance (Correspondence) D) Chronique (News) E) Notes et documents (Notes and documents) F) Bibliographie (Bibliography : Reviews or simple announcements of treatises) G) Bulletin bibliographique (Bibliographical bulletin : Content of the main mathematical journals)
The categorisation of the various contributions was not rigid and absolute ; thus sometimes the same type of contribution appeared under different headings. During these years there was little change in the way the material was organised (e.g. in names of the sections, the distribution of the subjects, etc.) in order to reect accumulated editorial experience or requests from readers, although the journal always maintained its general character of being exible, open and multipurpose. Major changes occurred, however, when the journal became the ofcial organ of ICMI in 1909, after which it carried reports of meetings of the commission as well as proposals for ICMI studies and their outcomes. We can obtain a better view of the nature of the journal from some detail of the contents published under the various headings.
11 ) I will use the acronym ICMI, but in the journal the commission was termed CIEM (Commission Internationale de lEnseignement Math matique). In Germany the acronym was e IMUK (Internationale mathematische Unterrichtskommission). In The Mathematical Gazette (1912) the commission was termed International Commission on Mathematical Teaching or International Commission on the Teaching of Mathematics. In the presentation of the questionnaire for the inquiry into the training of secondary teachers of mathematics [LEnseign. Math. 17 (1915), 129145] we nd International Commission on Mathematical Education in the English translation and Commissione internazionale dellinsegnamento matematico in the Italian one. Now ICMI stands for International Commission on Mathematical Instruction.
29
GENERAL ARTICLES. Section A, and sometimes Section B, contained the sort of articles one may expect in a standard mathematical journal. The general articles were classied into various sub-sections. Over time, one of these subsections, Organisation of teaching, became of increasing importance. After the birth of ICMI this sub-section contained the various business material of the commission. Other issues concerning ICMI appeared also in the sections Chronique and Notes et documents. Some of the articles, especially those by important mathematicians, had been published before elsewhere. Often papers roused discussion among the readers and reactions were published in the correspondence. Until 1903, besides the section General articles there was a section Etudes p dagogiques (Pedagogical studies), explicit evidence of the editors e wish to stress the links with mathematics teaching. MISCELLANIES. This section contained short articles of various types. In some years this part was attached to the correspondence section. CORRESPONDENCE. In the Correspondance section there were interesting contributions which, in some cases, could almost be considered as articles in their own right. From the point of view of interpreting the spirit of the journal this section is important in helping us to understand the readership and the kind of problems that interested them. We shall see later that it was a readers letter that inspired a major inquiry launched by the journal (into the methods of working of mathematicians). Letters are also important in showing that countries all over the world were reached by the journal, even if letters came predominantly from Europe. Major mathematicians gured among those who reacted to the articles published in the journal : L. E. J. Brouwer [13 (1911), 377380] who commented on a paper by G. Combebiac on the theory of measure, G. Peano [8 (1906), 315316] who reacted to the papers by E. Carvallo and V. Jamet on the convergence of series. NEWS. This section included all kinds of news about the mathematical community : announcements of death, awards, meetings, proposals of new national programmes, monuments to be erected, the activity of societies and academies. Thanks to this section the reader was able to participate in the life of the mathematical community. This section is also important for present historians for the variety of information it provides.
30
F. FURINGHETTI
NOTES AND DOCUMENTS. This section is devoted mainly to providing information on academic courses (topics developed, names of the professors). When the journal became the ofcial organ of ICMI this section contained the Proceedings of the works of the national sub-commissions. BIBLIOGRAPHY. This section contained short reviews of treatises. It is valuable for us today, both in giving us the reactions of contemporary readers to treatises that afterwards became famous, and also in providing us with information about forgotten works. BIBLIOGRAPHIC BULLETIN. This section provided information about new books (authors, publishers and town of publication, format, and price) and the tables of contents of some important mathematical journals, and proceedings of academies and societies. Journals of history of mathematics and elementary mathematics are also mentioned. The list of the publications is not exhaustive, but is a very rich source of information (see Appendix 1).
THEMES
AND AUTHORS
To give an idea of the themes treated in the journal, the contributions that appeared in the sections General articles and Pedagogical articles (excluding editorial notes published at the beginning of the volumes) have been classied in the Table below under the following categories 12 ) :
Alg Ana App Ari Ast Esp FMa Fou Geo His IC Inq Log Logi MaP Mec Algebra Analysis Applications Arithmetic Astronomy On the Esperanto language Financial mathematics Foundational themes Geometry History (including obituaries) ICMI Inquiries Logarithms Logic (including set theory) Mathematical Physics Mechanics Methodology in teaching Mathematical models Nomography National surveys of systems of instruction Org Organisation of mathematical instruction Phi Philosophical themes PhM Philosophy of mathematics Pro Probability Psy Psychology Soc Society ToN Number Theory Tri Trigonometry Var Various themes Vec Vectors Met Mod Nom NSu
12 ) My classication is made from a reading of the articles, without reference to the classication in Jahrbuch uber die Fortschritte der Mathematik, nor to the present classication in Mathematical Reviews or Zentralblatt. The journal used a classication different from mine; see the list published in volume 40 (19511954). On the difculties of classifying papers, and especially papers of the past, see [Furinghetti & Somaglia 1992].
31
CONTRIBUTIONS
THAT APPEARED IN THE JOURNAL, CLASSIFIED BY THEMES, AND THEIR DISTRIBUTION
Year 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 Alg 25 3 1 2 2 4 4 2 2 1 1 1 2 Ana 60 3 3 8 1 2 6 6 5 2 3 4 1 4 4 6 2 App 2 1 1 Ari 9 1 1 3 1 2 1 Ast 2 1 1 Esp 3 1 2 FMa 1 1 Fou 30 5 5 5 3 1 1 2 3 1 1 3 Geo 150 7 14 8 16 12 6 9 8 9 9 9 5 15 3 8 12 His 22 1 3 2 1 1 1 3 2 3 1 1 1 2 IC 16 1 2 4 2 3 1 3 19 1 1 3 5 7 2 Inq Log 2 1 1 Logi 9 2 3 1 2 1 2 2 MaP Mec 22 1 1 3 1 2 2 3 1 3 1 2 2 41 10 1 3 1 1 9 4 2 2 3 1 2 2 Met Mod 1 1 Nom 5 1 1 2 1 NSu 18 3 2 2 3 2 1 3 1 1 19 2 2 1 1 1 1 1 1 1 1 7 Org Phi 6 1 3 1 1 2 2 PhM Pro 4 1 1 1 1 Psy 1 1 Soc 1 1 ToN 16 1 1 5 2 2 1 4 10 1 1 2 1 1 1 2 1 Tri Var 2 1 1 Vec 4 1 1 2 Totals 504 35 36 38 34 26 26 41 37 36 25 32 24 32 23 26 33
Further information may be obtained from the list of materials published in the journal itself [40 (19511954), 124174]. In the period being considered we nd 212 authors of general or pedagogical articles. Some articles in the sub-section Organisation of teaching are anonymous. Collaboratively written articles are unusual. Authors contributing more than ve articles were : J. Andrade (France), A. Aubry (France), P. Barbarin (France), C. Berdell (France), V. Bobynin (Russia), C. Cailler e
32
F. FURINGHETTI
(Switzerland), G. Combebiac (France), L. Crelier (France), H. Fehr (Switzerland), G. Fonten (France), L. Godeaux (Belgium), C. A. Laisant (France), e H. Laurent (France), G. Loria (Italy), C. M ray (France), J. Richard (France), e E. Turri` re (France). e Contributions came from a number of countries 13 ) such as Algeria, Argentina, Austria and the regions of its old Empire, Belgium, Denmark, France, Germany, Greece, Italy, Japan, The Netherlands, Portugal, Rumania, Russia, Spain, Switzerland, UK, Ukraine, USA. Usually, at the end of an article, we nd the authors surname, together with a rst inital and the place from which the article was submitted ; some of these places are just a village so that it is not always easy to determine the country of origin. From the various data it would appear that the desire for internationalism was indeed realised, even if the core readership was in Europe. As regards contributions, France led the way, followed by Switzerland, and in total rather more than 50% of the authors wrote from these two countries. The majority of the contributions were on geometry. That this was at that time considered to be the backbone of the mathematical instruction at secondary level in many countries is shown by the many letters from readers discussing themes related to Euclidean geometry. Many of the articles on geometry concerned descriptive geometry and, in general, those aspects which have links with applications of interest to technical institutes and faculties. Among the authors contributing to this topic, J. Andrade published attempts at new approaches to geometry suitable for technical schools. A few of the articles dealt with non-Euclidean geometries. Often one feels that behind many articles lay the problem of answering such questions as the role of rigour and axiomatic methods in the teaching of mathematics. This subject is linked to the foundational debate, very much alive in those years. Indeed, following the birth of ICMI the debate on the place of foundations in mathematical instruction became the object of specic inquiries published in the journal. The themes of the reforms in mathematics teaching were treated in articles surveying mathematical systems of instruction and in articles that focused on specic parts of mathematics. For example, the teaching of analysis is discussed at a general level by Klein [1906], in line with the ideas expressed in the Meran Syllabus, and in articles discussing didactical problems of specic topics, such as the paper of Fehr [1905a] on the concept of function. In the following years the publication in the journal of the studies launched by ICMI contributed to raising interest in the theme of teaching calculus and analysis
13
) I refer to the modern names of the countries.
33
in many countries. An echo of this interest is to be found in national journals devoted to mathematics teaching, such as the British Mathematical Gazette and the Italian Bollettino della Mathesis. The theme of analysis is typical of the relationships between curricular innovations, mathematical research and the needs of society at the time (see [Bourlet 1910]). Also the interest shown by the journal in the book Trait de nomographie 14 ) by Maurice dOcagne e [1899] is evidence of the interest in the applications of mathematics and the teaching of mathematics to students in technical schools. At rst glance some teaching suggestions, such as those in [Sainte Lagu e 1910] on the use of squared paper, may appear naive or obvious to a modern reader, if one does not take into account the state of teaching methods of the time and the difculty of introducing new ideas, such as the graphing of functions (see [Gibson 1904 ; Brock & Price 1980]). The common mathematical interests of the two founders of the journal are illustrated by the attention paid to quaternions, the vectorial calculus and discussions about vectorial notations. In this regard we signal the article by J. S. Mackay [1905] on Peter Guthrie Tait. The journal was open to a wide range of contributions and ideas even if, in some cases, the editors pointed out that they did not share the contributors opinions. Among the authors there were various important characters in the world of mathematics education and of history of mathematics (Loria, Smith, etc.), famous mathematicians (E. Borel, C. Bourlet, L. E. J. Brouwer, E. Czuber, G. Darboux, F. Enriques, M. Frechet, Z. G. de Galdeano, J. Hadamard, D. Hilbert, F. Klein, H. Lebesgue, B. Levi, C. M ray, P. Painlev , H. Poincar , e e e F. Gomes Teixeira, H. Weyl, etc.), philosophers such as L. Couturat, and a cohort of secondary and university teachers now forgotten, but very active at the time in contributing to the international debate.
THE
INQUIRY ON THE METHOD OF WORKING OF MATHEMATICIANS
The editorial line of the journal ranged between two poles, one specically referring to mathematics and the other concerned with pedagogical aspects linked to mathematics teaching. Psychology was a link between these two poles. In writing on mathematical topics some authors hinted at psychological
14 ) As discussed in dOcagnes book, nomography is a theory which permits the graphical representation of mathematical laws dened by equations of any number of variables (see the review of the book in LEnseign. Math. 1 (1899), 368370). A passage of dOcagnes book is translated in [Smith 1929].
34
F. FURINGHETTI
issues. For example, the two articles [Baron 1903 ; Bonola 1903] show on the one hand the attention paid to strands of contemporary mathematical research (foundational studies), and on the other hand the importance of explaining mathematical facts through factors exterior to mathematics. On the reports being published on foundational studies in geometry, Bonola wrote that they
ont non seulement elargi le domaine de la G om trie non-euclidienne, mais e e ils ont encore contraint les esprits bien organis s a suivre les edices vari s et e ` e e admirables qui ont et inspir s par la renaissance scientique, caract ristique e e ` du XIXe si` cle, et a se rendre compte des connexions qui existent entre les e e branches les plus elev es des math matiques et dint ressants probl` mes de e e e Psycho-physiologie. 15 ) [Bonola 1903, 317]
At that time, themes of psychology, such as the necessary conditions for creativity and invention, and their links with mathematical themes such as axiomatisation, rigour and intuition, were debated hotly in the mathematical community (see [Mannheim 1909 ; Poincar 1908]). e This context may explain why the inquiry into the methods of mathematicians was promoted by the journal. This was rst inspired by a letter from Ed. Maillet [Correspondance, 3 (1901), 5859]. He wrote that it would be interesting for young researchers beginning their profession to have information on : work and research methods, general hygienic rules useful to their intellectual work, how to read papers, etc. In the same year [Chronique, 3 (1901), 128 ; 219220] the editors acknowledged the great interest roused by the idea and launched the project of a questionnaire, which had to respect condentiality and was to be more than a mere curiosity. Many readers responded to this note and sent questions to be included in the questionnaire 16 ). The rst part of the questionnaire was published in 1902 [4, 208211] and the nal part in 1904 [6, 376378]. The questionnaire was sent to the subscribers of the journal and also distributed to the participants at the International Congress of Mathematicians at Heidelberg in 1904 and those attending the congress of Saint-Louis [LEnseign. Math. 6 (1904), 481]. In Heidelberg, Fehr [1905b] presented a communication on the questionnaire.
15 ) have not only extended the domain of non-Euclidean geometry, but have also obliged well organised minds to pay regard to various and admirable constructions, inspired by the scientic renaissance characteristic of the 19th century, and to take account of the links between the highest branches of mathematics and interesting problems of psycho-physiology. 16 ) At this time Ed. Maillet was one of the editors of the journal LInterm diaire des e Math maticiens. In 1902 (t. IX) that journal published two questions [question numbers : 2446 and e 2447, 263264] on the circumstances attending mathematical creation. Answers to these questions were published in the same volume [339343].
35
Copies of the questionnaire were sent to other mathematicians and to those who requested it. In 1905 [7, 239] the editors thanked the scientic journals which had published the questions. The aim of the enterprise was twofold : on the one hand to collect advice useful to researchers in mathematics, and on the other hand to contribute to research in the eld of psychology of professions. The questionnaire was prepared and analysed by Fehr and two psychologists at the University of Geneva, Edouard Clapar` de (18731940) and Th odore Flournoy (1854 e e 1920). The languages for the answers were restricted to English, Esperanto, French, German, and Italian. There were 30 questions in all and the names of responders were published only with their explicit agreement. Clapar` de, Fehr e and Flournoy published the results with comments in 1905, 1906, 1907, and 1908. There was also a report of the inquiry published in book form [Fehr et al. 1908]. The analysis of the results was based on the answers of more than 100 mathematicians and on a few historical notes on the lives of previous mathematicians. The journal published statistical data, comments and a few sentences from the responders. Many of the published statements carried the names of the writers, so that we know that among the responders were L. Boltzmann, M. Cantor, J. Coolidge, L. E. Dickson, H. Fehr, Z. G. de Galdeano, C. A. Laisant, C. M ray, G. Oltramare and J. W. Young. With e these statements the nationality of the responder was also indicated. The answers arrived from various countries. Comments to the questions, by G. Loria [8 (1906), 383385, questions 6 and 9] and V. Bobynin [9 (1907), 135141 ; 389396, questions 4 and 5] were published 17 ). This inquiry is not mere folklore ; it provides material for studies in mathematics education, epistemology, psychology and sociology. Also it constitutes an early example of making explicit the feeling mathematicians have about the nature of their discipline, of their work and of their being mathematicians. This willingness of mathematicians to be more open about their feelings and how they worked was shown in later years by a number of famous works (by Hadamard, Hardy, Poincar , etc.). As concerns the e results, Fehr himself [1908] acknowledged that it was difcult to draw general conclusions. As a matter of fact, the importance of this enterprise was in identifying themes which would have important developments in the following years. Even if the intentions of the designers of the questionnaire were mainly directed towards mathematical research, the questionnaire may have had an
17
) Appendix 2 contains the list of notes and articles dealing with this inquiry.
36
F. FURINGHETTI
impact in the world of school, by introducing elements pertaining to the psychological domain in the reection on students mathematical performances. For example, we can cite the report of the German national commission Psychologie und mathematischer Unterricht by D. Katz [1913]. The review [Brandenberger 1914] reports that it was divided into three parts :
psychology and mathematics teaching (the childs development of the
representation of number and the conception of space, the methods of working of mathematicians in line with the inquiry published in LEnseignement Math matique, the psychology of great mental reckoners, e teaching to low-attainers, etc.),
psychology of technical and artistic drawing, psychology and pedagogy in teacher education.
This report therefore illustrates the impact which the LEnseignement Math e matique inquiry had on the world of mathematics education. In the preface to the posthumous article by Mannheim [1909], the editors reported that the inquiry had had unexpected side effects, for example Mannheims and Poincar s papers. As a matter of fact, the themes developed e through this inquiry were close to the interests of a prominent member of the Committee of supporters, Henri Poincar , who contributed four articles e to the journal. The paper [Poincar 1908], which is the text of a lecture e delivered at the Institut g n ral psychologique, may be seen as a bridge e e between the world of mathematicians and that of psychologists. It shows that this prominent mathematician, as well as Mannheim in his paper of 1909, was in agreement with the ideas of the editors of LEnseignement Math matique. e At the beginning of the paper Poincar wrote : e
La gen` se de linvention math matique est un probl` me qui doit inspirer le e e e plus vif int r t au psychologue. Cest lacte dans lequel lesprit humain semble ee le moins emprunter au monde ext rieur, o` il nagit ou ne parat agir que par e u lui-m me et sur lui-m me, de sorte quen etudiant le processus de la pens e e e e g om trique, cest ce quil y a de plus essentiel dans lesprit humain que nous e e pouvons esp rer atteindre. e On la compris depuis longtemps, et, il y a quelques mois, une revue intitul e LEnseignement Math matique, et dirig e par MM. Laisant et Fehr, a e e e entrepris une enqu te sur les habitudes desprit et les m thodes de travail des e e diff rents math maticiens. Javais arr t les principaux traits de ma conf rence e e ee e e quand les r sultats de cette enqu te ont et publi s ; je nai donc gu` re pu e e e e ` les utiliser. Je me bornerai a dire que la majorit des t moignages conrment e e mes conclusions ; je ne dis pas lunanimit , car, quand on consulte le suffrage e universel, on ne peut se atter de r unir lunanimit . 18 ) [Poincar 1908, 357] e e e
18
) The genesis of mathematical creation is a problem which ought to inspire the intense
37
The Swiss environment was important in encouraging this kind of interest by the journal and one may say that the roots of future developments of the Geneva school of psychology 19 ) may be found in those years. To grasp the particular atmosphere of the period it is interesting to consider the review by Clapar` de [LEnseign. Math. 14 (1912), 8182] of the book e Henri Poincar (by Dr. Toulouse, Flammarion, Paris). This book presents the e results of a study on Poincar s way of reasoning and a comparison with the e results obtained from similar studies on the novelist Emile Zola. Clapar` de e reports that the conclusions of the author are that Zolas intelligence is
consciente, logique, m thodique, paraissant faite pour la d duction math e e e matique []. Au contraire lactivit mentale de M. Poincar est spontan e, e e e peu consciente, plus proche du r ve que de la d marche rationnelle, et semblait e e surtout apte aux uvres de pure imagination []. 20 )
THE
DAWN OF
ICMI
The international vocation of LEnseignement Math matique was clearly e established in the rst volume of the journal with the articles on mathematics teaching in Spain by Z. G. de Galdeano and on the history of mathematics teaching in Russia by V. Bobynin. In successive issues similar surveys of other countries were published. The readers were gradually invited to consider the importance of comparative studies on different systems of mathematical development (intended in the broad sense of including both research and instruction).
interest of the psychologist. It is the activity in which the human mind appears to take least from the outside world, where it acts or seems to act only of itself and on itself, so that in studying the process of geometric thinking we may hope to reach what is most essential in mans mind. This has long been appreciated, and some months ago a journal called LEnseignement Math matique, edited by Laisant and Fehr, began an investigation of the mental habits and e methods of work of different mathematicians. I had already dened the main outlines of my lecture when the results of that inquiry were published; I have not therefore been able to make much use of them. I shall conne myself to saying that the majority of witnesses conrm my conclusions; I do not say all, for when the appeal is to universal suffrage one can hardly expect to obtain unanimity.
19 ) Jean Piaget never missed an opportunity to mention the ideas of Clapar` de (see, for e example, [Beth & Piaget 1961, 213214]). He himself had been called by Edouard Clapar` de (in e 1921) to teach at the J.-J. Rousseau Institute in Geneva. As a matter of fact there is a discussion of the Inquiry with reference to Poincar and Hadamards conceptions in [Beth & Piaget 1961, e 26, 9699]. We may add that Piagets support was very useful when the second series of the journal was launched in 1955, and he even became a member of the editorial board. 20 ) conscious, logical, methodical, appearing suitable for mathematical deduction []. On the contrary, the mental activity of Poincar is spontaneous, hardly conscious, closer to dreams than e to the rational way, and seems most of all suitable for works of pure imagination [].
38
F. FURINGHETTI
These ideas became more concrete in the note [La R daction 1905b] e published in the seventh year, where the editors call for the opinions of readers on the following questions : 1. What progress needs to be achieved in the organisation of the teaching of pure mathematics ? 2. What should be the role of tertiary instruction in teacher education ? 3. How should teaching be organised in such a way as to meet the requirements of pure and applied sciences better than in the past. As indicated in the note, an important antecedent to this initiative lay in the resolution passed by the third International Congress of Mathematicians held in Heidelberg as reported on p. 53 of the Proceedings [Krazer 1905] and in a note by La R daction [1905a]. This resolution stated that the teaching of e mathematics at the tertiary level had to be in accordance with the importance of technical sciences in the various countries. So it was stressed that there was a need for new professorial appointments, suitable libraries, collections of models, laboratories for drawing and practical work, etc. J. Andrade, E. Borel, G. Loria, F. Marotte and D. E. Smith reacted to the questions with notes published in the journal [LEnseignement Math matique e 1905], in which the key issues for reform were stressed. The discussion was enlarged to include the theme of mathematics teaching at the secondary level. Teacher education was the natural link between these two levels. Loria [1905] advocated the democratisation of some theories that in those years were considered advanced (e.g. analytical geometry). Smith [1905] proposed more international co-operation and the creation of a commission to be appointed in an international conference for the study of instructional problems globally. He stated that both the journal and the commission were the best means to improve mathematics teaching. Smith was relying on the community of mathematicians, but at the same time he was quite critical about it in remarking that the papers presented in the didactic section of the International Congress of Mathematicians in Heidelberg were discussions of mathematical details rather than of general problems of teaching. The discussion on the need for educational reforms in LEnseignement Math matique was the stimulus for the idea of a commission appointed by an e international congress, which had its culmination in the establishment of ICMI during the fourth International Congress of Mathematicians in Rome in 1908 (see [Howson 1984 ; Schubring 2003]). Fehr was appointed as the secretarygeneral of ICMI, with Felix Klein as president and Alfred George Greenhill as vice-president. The following year LEnseignement Math matique became e
39
the ofcial organ of ICMI. In particular, through the journal, ICMI launched its early studies on various themes of mathematics teaching (cf. [Schubring 2003]). The reports of the discussions on these themes were published in the proceedings of important conferences organised by ICMI : 1911 in Milan, 1912 during the International Congress of Mathematicians in Cambridge, and 1914 in Paris (see Appendix 3). Also the journal published regular news about the Sub-Commissions founded in the various countries. Pamphlets and volumes on the works of these sub-commissions were also published (see [Fehr 1920]). Together they constitute important documentation on the state of mathematical instruction at that time.
CONCLUSION The journal LEnseignement Math matique played a unique role in the e mathematical community (of research and of instruction) in the period I have considered. It was a bridge between some aspects of mathematical research, the world of school and society. It was also a forum for discussing problems of mathematical instruction and education. In addition it was a bridge between mathematics and other disciplines, such as philosophy, psychology, and technology. The journal contributed to redirecting the attention of educators away from the core of traditional mathematical instruction (Euclidean geometry) towards new themes (analytical geometry, foundational studies, and analysis) linked to technical applications or to research being developed in those years. This contributed to raising awareness in a number of countries of the need for a discussion of the nature of their mathematical programmes. For example, in Italy the scientic lyceum, whose curriculum encompasses a strong mathematical programme, was created under the inuence of the debate on the teaching of calculus and the study launched by ICMI on this subject (see [Furinghetti 2001]). Some of the ideas discussed in LEnseignement Math matique were already e gaining ground in the scientic environment ; the journal had the merit of being an efcient promoter of these ideas and of providing the fertile soil in which the ideas could grow.
40
F. FURINGHETTI
APPENDIX 1 JOURNALS
OF LISTED 21 ) IN THE SECTION
BULLETIN
YEARS
BIBLIOGRAPHIQUE
LENSEIGNEMENT MATHEMATIQUE
1899, 1900, 1901
(WITH
THE TOWN OF PUBLICATION THE FIRST TIME THE JOURNAL IS QUOTED)
Acta Mathematica, Stockholm (Sweden) American Journal of mathematics, Baltimore (USA) Annales de la Facult des sciences de lUniversit de Toulouse, Paris, Toulouse (France) e e Annali di Matematica pura ed applicata, Milan (Italy) Annals of Mathematics, Cambridge (USA) Archiv der Mathematik und Physik, Leipzig (Germany) Atti della Reale Accademia dei Lincei, Rendiconti, Rome (Italy) Bibliotheca mathematica, Zeitschrift fur Geschichte der mathematischen Wissenschaften, Leipzig (Germany) Bolletino di Bibliograa e Storia delle Scienze mathematiche, Turin (Italy) Bulletin astronomique, Paris (France) Bulletin de lEnseignement technique, Paris (France) Bulletin de la Soci t math matique de France, Paris (France) ee e Bulletin de la Soci t Philomathique de Paris, Paris (France) ee Bulletin de math matiques sp ciales, Paris (France) e e Bulletin des Sciences math matiques, Paris (France) e e Bulletin des sciences math matiques et physiques el mentaires, Paris (France) e Comptes rendus des s ances de lAcad mie des sciences, Paris (France) e e Educational Review, Rochway, New York (USA) El Progreso matematico, Zaragoza (Spain) Giornale di matematiche di Battaglini, Naples (Italy) Il Bollettino di matematiche e di scienze siche e naturali, Bologna (Italy) Il nuovo Cimento, Pisa (Italy) Il Pitagora, Palermo (Italy) Jahrbuch ueber die Fortschritte der Mathematik, Berlin (Germany) Jahresbericht der Deutschen Mathematiker Vereinigung, Leipzig (Germany) Jornal de Sciencias mathematicas e Astronomicas, Coimbra (Portugal) Jornal de sciencias mathematicas, physicas e naturs, Lisboa (Portugal) Journal de lEcole Polytechnique, Paris (France)
21
) Spelling and capitalization as they appeared the rst time each journal was mentioned.
41
e Journal de math matiques el mentaires, Paris (France) e e Journal de math matiques el mentaires ; journal de math matiques sp ciales, Paris e e e (France) Journal de Math matiques pures et appliqu es, Paris (France) e e Journal de math matiques sp ciales, Paris (France) e e Journal f r die reine und angewandte Mathematik, Berlin (Germany) u LEducation math matique, Paris (France) e LEnseignement chr tien, Paris (France) e LInterm diaire des Math maticiens, Paris (France) e e Le Matematiche pure ed applicate, Citt` di Castello (Italy) a Mathematische Annalen, Leipzig (Germany) ` Mathesis, recueil math matique a lusage des ecoles sp ciales et des etablissements e e dinstruction moyenne, Gand (Belgium), Paris (France) Monatschefte f r Mathematik und Physik, Vienna (Austria) u Nieuw Archief voor Wiskunde, Amsterdam (Netherlands) Nouvelles Annales de math matiques, Paris (France) e Nyt Tidsskrift for matematik, Copenhagen (Denmark) Paedagogisches Archiv, Leipzig (Germany) Periodico di Matematica (per linsegnamento secondario), Leghorn (Italy) Publications de la section des sciences math matiques et naturelles de lUniversit e e dUpsall, Uppsala (Sweden) Rendiconti del Circolo matematico di Palermo, Palermo (Italy) Revista de Ciencias, Lima (Peru) Revista trimestral de matematicas, Zaragoza (Spain) Revue de Math matiques (Rivista di Matematica), Turin (Italy) e Revue de math matiques sp ciales, Paris (France) e e e Revue de Physique exp rimentale et de Math matiques el mentaires, Odessa (Russia) e e Revue g n rale des sciences pures et appliqu es, Paris (France) e e e Revue scientique, Paris (France) Revue semestrielle des publications math matiques, Amsterdam (Netherlands), Edine burgh (UK), Leipzig (Germany), London (UK), Paris (France) Schweizerische Paedagogische Zeitschrift, Zurich (Switzerland) Sciences physico-math matiques, Moscow (Russia) e Supplemento al periodico di Matematica, Leghorn (Italy) The american mathematical Monthly, Springeld (USA) The Mathematical Gazette, London (UK) Transactions of the American Mathematical Society, Lancaster, New York (USA) Unterrichtsbl tter f r Mathematik und Naturwissenschaften, Berlin (Germany) a u Wiadomosci Matematyczne, Warsaw (Poland) Wiskundige Opgaven, Amsterdam (Netherlands) Zeitschrift f r das Realschulwesen, Vienna (Austria) u Zeitschrift f r lateinlose hoehere Schulen, Leipzig (Germany) u Zeitschrift f r Mathematik und Physik, Leipzig (Germany) u Zeitschrift f r mathematischen und naturwissenschaftlichen Unterricht, Leipzig (Geru many)
42
F. FURINGHETTI
APPENDIX 2 THE
INQUIRY ON THE METHOD OF WORKING OF MATHEMATICIANS
QUESTIONS 22 )
Text of the rst questionnaire : 4 (1902), 208211 Text of the second questionnaire : 6 (1904), 376378
ANALYSIS
OF THE RESULTS
Question 1a : H. Fehr, 7 (1905), 387395 Question 1b : T. Flournoy, 7 (1905), 473478 Questions 2 and 3 : H. Fehr, 8 (1906), 4348 Questions 4 and 5 : H. Fehr, 8 (1906), 217225 Questions 69 : T. Flournoy, 8 (1906), 293310 Questions 1013 : H. Fehr, 8 (1906), 463475 Questions 1417 : H. Fehr, 9 (1907), 123128 Questions 18 and 20 : T. Flournoy, 9 (1907), 128135 Question 19 : T. Flournoy, 9 (1907), 204217 Question 21 : H. Fehr, 9 (1907), 306312 Questions 22 and 23 : E. Clapar` de, 9 (1907), 473479 e Questions 2430 : E. Clapar` de, 10 (1908), 152172 e
NOTES ABOUT THE INQUIRY ON THE METHOD OF WORKING OF MATHEMATICIANS

(1901), 5859 (Maillet, Ed., Correspondance) ; 128 ; 219220 (1904), 481 (1905), 6364 ; 239240 (1906), 383385 : letter of G. Loria on questions 69 (1907), 135141 : reections of V. Bobynin on the answers to questions 4 and 5 9 (1907), 389396 : reections of V. Bobynin on the answers to questions 1113 3 6 7 8 9
22 ) For an English translation of the questions see Appendix I in Hadamards book [Hadamard 1945, 136141].
43
APPENDIX 3 INQUIRIES
ON THE TEACHING OF MATHEMATICS LAUNCHED BY
ICMI
13 (1911) The questions discussed in the conference of the Commission in Milan (1821 September 1911) are published on pp. 122 ; 443446. THEMES : A) Systematic exposition of mathematics (axioms, rigour, etc.) in secondary schools ; the fusion of different branches of mathematics in secondary schools. B) The teaching of mathematics to university students of physics and of the natural sciences. The Proceedings of the Milan conference with reports about these questions can be found in the same volume [437511]. 14 (1912) The questions discussed in the conference held during the International Mathematical Congress (Cambridge, UK, 2128 August 1912) are published on pp. 39 ; 132135 ; 220 ; 299. THEMES : A) Intuition and experimental evidence in secondary schools. B) Mathematics for university students of physics. The Proceedings of the Cambridge conference with reports about these questions can be found in the same volume [441537]. 15 (1913) The questions to be discussed in preparatory work for the conference to be held in Paris (1914) are reported on pp. 243 ; 414 ; 487. The translations of these questions into German, English, and Italian can be found on pp. 394412. THEMES : A) Results of the introduction of calculus in secondary schools. B) Mathematics teaching for the technical professions in higher educational institutions. 16 (1914) The questions for the Paris conference are published on p. 54. The Proceedings of the Paris conference (14 April 1914) with reports about these questions are published on pp. 165226 and 245356.
44
F. FURINGHETTI
REFERENCES ` ANDRADE, J. Opinion de Jules Andrade sur les r formes a accomplir dans lenseignement e des math matiques. LEnseign. Math. 7 (1905), 462469. e ANONYMOUS. Henri Fehr (18701954). Sa vie et son uvre. LEnseign. Math. (2) 1 (1955), 410 (with a portrait). BARON, R. Philologues et psychologues en face du probl` me des parall` les. LEnseign. e e Math. 5 (1903), 279287. BEHNKE, H. Commission internationale de lenseignement math matique. Sa particie pation au Congr` s dAmsterdam 1954. LEnseign. Math. 40 (195154), 7274. e BETH, E. W. et J. PIAGET. Epist mologie math matique et psychologie. P.U.F., Paris, e e 1961. BONOLA, R. A propos dun r cent expos des principes de la g om trie non-euclidienne. e e e e LEnseign. Math. 5 (1903), 317325. ` BOREL, E. Opinion dEmile Borel sur les r formes a accomplir dans lenseignement e des math matiques. LEnseign. Math. 7 (1905), 386387. e BOURLET, C. La p n tration r ciproque des enseignements des math matiques pures et e e e e des math matiques appliqu es dans lenseignement secondaire. LEnseign. Math. e e 12 (1910), 372387. BRANDENBERGER, C. Psychologie et enseignement math matique. LEnseign. Math. 16 e (1914), 141142. BROCK, W. H. and M. H. PRICE. Squared paper in the nineteenth century : instrument of science and engineering, and symbol of reform in mathematical education. Educational Studies in Mathematics 11 (1980), 365381. BUHL, A. Charles-Ange Laisant (18411920). LEnseign. Math. 21 (1920), 7280 (with a portrait). FEHR, H. [1905a] La notion de fonction dans lenseignement math matique des ecoles e moyennes. LEnseign. Math. 7 (1905), 177187. [1905b] Lenqu te de LEnseignement Math matique sur la m thode de e e e travail des math maticiens. In : A. Krazer (ed.), Verhandlungen des dritten e internationalen Mathematiker-Kongresses, 603607. B. G. Teubner, Leipzig, 1905. Note nale. LEnseign. Math. 10 (1908), 171172. Commission internationale de lenseignement math matique. LEnseign. Math. e 16 (1914), 477478. ` La Commission internationale de lenseignement math matique de 1908 a 1920. e Compte rendu sommaire suivi de la liste compl` te des travaux publi s par la e e Commission et les Sous-commissions nationales. LEnseign. Math. 21 (1920), 305342. A. Buhl (18781949). LEnseign. Math. 39 (19421950), 68. ` FEHR, H. (with the collaboration of T. FLOURNOY and E. CLAPAREDE). Enqu te de e LEnseignement Math matique sur la m thode de travail des math maticiens. e e e Gauthier-Villars, Paris Georg & Co., Gen` ve, 1908. e FRECHET, M. Pri la funkcia ekvacio f (x y) f (x) f (y) . LEnseign. Math. 15 (1913), 390393 ; 16 (1914), 136.

45
FRIEDELMEYER, J.-P. La cr ation des premi` res revues de math matiques et la distinction e e e math matiques pures, math matiques appliqu es. In : Actes de la 6e Universit e e e e d t interdisciplinaire sur lhistoire des math matiques (Besancon, 1995), 215 ee e 236. Editions et diffusion IREM de Besancon, 1996. FURINGHETTI, F. Il Bollettino della Mathesis dal 1909 al 1920 : pulsioni tra temi didattici internazionali e nazionali. PRISTEM/Storia. Note di Matematica, Storia, Cultura 5 (2001), 3158. FURINGHETTI, F. e A. SOMAGLIA. Giornalismo matematico a carattere elementare nella seconda met` dellOttocento. Linsegnamento della Matematica e delle Scienze a Integrate 15 (1992), 815852. GIBSON, G. A. An Elementary Treatise on Graphs. Macmillan and Co, LondonNew York, 1904. HADAMARD, J. An Essay on the Psychology of Invention in the Mathematical Field. Princeton University Press, 1945 (also Dover, New York, 1954). HOWSON, A. G. Seventy-ve years of the International Commission on Mathematical Instruction. Educational Studies in Mathematics 15 (1984), 7593. KATZ, D. Psychologie und mathematischer Unterricht. In : Abhandlungen uber den mathem. Unterricht in Deutschland, Band III, Heft 8. B. G. Teubner, Leipzig, 1913. KLEIN, F. De lenseignement des sciences math matiques et physiques dans les e universit s et hautes ecoles techniques. LEnseign. Math. 8 (1906), 525. e KRAZER, A. (ed.) Verhandlungen des dritten internationalen Mathematiker-Kongresses. B. G. Teubner, Leipzig, 1905. ` LA REDACTION. [1905a] Lenseignement des math matiques a lUniversit . LEnseign. e e Math. 7 (1905), 238. ` [1905b] Note de la R daction sur les r formes a accomplir dans lenseignement e e des math matiques. LEnseign. Math. 7 (1905), 382383. e ` LENSEIGNEMENT MATHEMATIQUE. R formes a accomplir dans lenseignement des e math matiques. LEnseign. Math. 7 (1905), 382387, 462472. e LES DIRECTEURS. LEnseignement Math matique. LEnseign. Math. 1 (1899), 15. e ` LORIA, G. Opinion de Gino Loria sur les r formes a accomplir dans lenseignement e des math matiques. LEnseign. Math. 7 (1905), 383386. e MACKAY, J.-S. Peter Guthrie Tait. LEnseign. Math. 7 (1905), 510. MANNHEIM, A. Linvention math matique. LEnseign. Math. 11 (1909), 161167. e ` MAROTTE, F. Opinion de F. Marotte sur les r formes a accomplir dans lenseignement e des math matiques. LEnseign. Math. 7 (1905), 471472. e MERAY, C. L esperanto , langue auxiliaire articielle de M. le Dr Zamenhof, ouvrant ` les plus larges perspectives a la litt rature scientique internationale. LEnseign. e Math. 2 (1900), 265293. Correspondances internationales en esperanto. LEnseign. Math. 3 (1901), 3146. DOCAGNE, M. Trait de nomographie. Gauthier-Villars, Paris, 1899. e ORTIZ, E. L. Quaternions abroad : some remarks on their impact in France. Acta Histori Rerum Naturalium necnon Technicarum. New series, 3 (1999), 295302. Proyectos de cambio cientco y proyectos de cambio poltico en la tercera Rep blica : el caso de la teora de los cuaterniones. Revista Brasileira de Historia u da Matem tica 1 (2001), 6685. Also in : El valor de las matematicas (edited a by A. Duran & J. Ferreiros), 141164. Sevilla, 2001.
46
F. FURINGHETTI
PARSHALL HUNGER, K. Mathematics in national contexts (18751900) : An international overview. In : S. D. Chatterji (ed.), Proceedings of the International Congress of Mathematicians (Zurich, 1994), vol. 2, 15811591. Birkh user, BaselBoston a Berlin, 1995. ` PASCAL, J. Les d put s bretons de 1789 a 1983. P.U.F., Paris, 1983. e e POINCARE, H. Linvention math matique. LEnseign. Math. 10 (1908), 357371. e DE RHAM, G. Henri Fehr (18701954). Verhandlungen der schweizerischen naturforschenden Gesellschaft 135 (1955), 334339 (with a portrait). LEnseignement Math matique Revue internationale et la Commission Internae tionale de lEnseignement Math matique (CIEM). ICMI Bulletin 7 (April 1976). e ROERO, S. I matematici e la lingua internazionale. Bollettino UMI 8, II-A (1999), 159182. RUFFET, J. Henri Fehr. Elemente der Mathematik 10 (1955), 14. SAINTE LAGUE, A. Note sur les usages du papier quadrill . LEnseign. Math. 12 (1910), e 517. SAUVAGE, J.-P. Charles-Ange Laisant, d put et math maticien. Histoire, m moires e e e e locales, d partementales, r gionales 2 (automne 1994). e e SCHUBRING, G. La r forme de lenseignement des math matiques en Allemagne dans e e les ann es 19001914 et son r le dynamique dans le mouvement international e o de r forme. In : B. Belhoste, H. Gispert et N. Hulin (eds), Les sciences au lyc e. e e ` Un si` cle de r formes des math matiques et de la physique en France et a e e e l tranger, 235248. Vuibert INRP, Paris, 1996. e [2003] LEnseignement Math matique and the rst International Commission e (IMUK) : the emergence of international communication and cooperation. This volume, 4765. ` SMITH, D. E. Opinion de David-Eugene Smith sur les r formes a accomplir dans e lenseignement des math matiques. LEnseign. Math. 7 (1905), 469471. e A Source Book in Mathematics. McGraw-Hill, New York, 1929 (also Dover, New York, 1959). VASSILIEF, A. Les id es dAuguste Comte sur la philosophie des math matiques. e e LEnseign. Math. 2 (1900), 157172. WHITEHEAD, A. N. The principles of mathematics in relation to elementary teaching. LEnseign. Math. 15 (1913), 105112.
LENSEIGNEMENT MATHEMATIQUE AND THE FIRST INTERNATIONAL COMMISSION (IMUK) : THE EMERGENCE OF INTERNATIONAL COMMUNICATION AND COOPERATION LEnseignement Math matique et la premi` re Commission internationale : e e naissance de communications et de coop rations internationales e par Gert SCHUBRING
Le centenaire de la revue LEnseignement Math matique r v` le le contraste existant e e e au d but du si` cle dernier entre les niveaux de d veloppement de la communication e e e internationale, dune part en math matiques et dautre part dans leur enseignement. e Depuis longtemps, en effet, il y avait chez les math maticiens tant une communication e directe que des revues math matiques, soutenues par des auteurs et un public e internationaux, ou encore des rencontres, tel le deuxi` me Congr` s international des e e math maticiens de Paris qui permit de faire un bilan des avanc es du domaine tel quil e e se pr sentait en 1900. Mais pour ce qui est de l ducation math matique, un pareil e e e niveau de communication parmi les enseignants na pu etre etabli quavec la fondation de la nouvelle revue, en 1899. A lexamen des grands pays europ ens, on voit que la situation de lenseignement e des math matiques y etait tellement diverse pendant le XIXe si` cle quun tel manque e e ` de communication internationale est tout a fait compr hensible. M me en France, e e e lenseignement des math matiques pour la grande majorit des el` ves dans les e e lyc es noccupait quune place mineure, de sorte quune professionnalisation des e enseignants de math matiques pouvait difcilement avoir lieu. En Italie et en Angleterre, e lenseignement etait caract ris par une conception strictement classique du savoir e e ` enseign , pr conisant ainsi Euclide a la fois comme manuel et comme m thodologie e e e pour les math matiques. e Avec la fondation de la revue LEnseignement Math matique, la situation commenca e ` a changer : les rapports publi s d` s lors sur l tat de lenseignement des math matiques e e e e dans divers pays rent natre le d sir d tablir une coop ration internationale. Le e e e
48
G. SCHUBRING
` premier a proposer la cr ation dune commission internationale pour promouvoir une e telle coop ration fut lAm ricain David Eugene Smith en 1905. e e La Commission internationale de lenseignement math matique (CIEM/ICMI), cr ee e e ` lors du Congr` s international des math maticiens suivant, tenu en 1908 a Rome, e e prit tr` s au s rieux sa mission de transformer les communications naissantes en une e e vraie coop ration internationale. An de promouvoir des r formes dans lenseignement e e math matique par lactualisation des programmes et des m thodes denseignement, la e e ` Commission elargit nettement le mandat restreint d ni par le Congr` s en invitant a e e participer des repr sentants de pays plus nombreux quinitialement pr vu (et de tous e e les continents). La Commission ne se contenta pas des seules ecoles secondaires, ` e mais r solut de sint resser a tous les types d coles, depuis l cole el mentaire e e e e jusqu` lenseignement sup rieur, en passant par les ecoles d ducation g n rale et a e e e e les ecoles professionnelles ou techniques. M me si la grande majorit des membres e e e de la Commission appartenait au secteur post-secondaire et aurait de ce fait et naturellement li e principalement aux ecoles secondaires, ils soutinrent activement e cette politique universaliste du Comit Central. e Cest seulement dans le choix des sujets pour des etudes comparatives visant ` a promouvoir lesprit des r formes que lon peut remarquer un biais r sultant des e e orientations professionnelles des membres : parmi les huit sujets retenus, trois seulement ` concernaient le mandat ofciel, a savoir les ecoles secondaires, les cinq autres portant soit sur la transition du secondaire au sup rieur, soit exclusivement sur les probl` mes e e du secteur post-secondaire. On pr sente dans ce texte un r sum des d bats et des r sultats concernant certaines e e e e e de ces etudes comparatives. En conclusion, on compare le d veloppement international e de lenseignement math matique aujourdhui avec la situation dil y a cent ans. e
LENSEIGNEMENT MATHEMATIQUE AND THE FIRST INTERNATIONAL COMMISSION (IMUK) : THE EMERGENCE OF INTERNATIONAL COMMUNICATION AND COOPERATION by Gert SCHUBRING
This World Mathematical Year 2000 provides us with an excellent opportunity for illustrating the specic situation of mathematics education. As we commemorate the International Congress of Mathematicians (ICM) in Paris in 1900, we can see how rmly international communication had been established among mathematicians at the time of that second international Congress. And Hilberts famous speech set the agenda for mathematical research of a large part of the new 20th century. But what was the state of international communication in mathematical education ? We should note that communication between different countries regarding the teaching of mathematics had been practically non-existent up to the end of the 19th century. While mathematicians in Europe had been communicating for centuries and journals accepting contributions from different countries had existed since the early 19th century international congresses in 1897 establishing a new, though logical development , founding the journal LEnseignement Math e matique was a bold endeavour since the necessary basis for international communication on matters of education had hardly been established at that time. This lack of international communication with regard to mathematical education was basically due to the fact that differences were profound, even between the countries of Western Europe. These differences concerned the contents of instruction as well as the epistemology of school mathematics and the methodology of teaching. A clear indicator is the amount of teaching time allotted to mathematics, as compared to that of competing school disciplines. Thus, both the number of weekly hours and the status of mathematics teachers were indicators for the social status enjoyed by mathematics as a teaching
50
G. SCHUBRING
subject in general education. I shall give a few examples to illustrate the considerable differences between European countries. Let us begin with FRANCE, a country which was generally believed to have established a strong position of mathematics in schools since the scienceminded reforms arising from the French Revolution. Actually, the inuence of these reforms declined decisively after the Restoration, leaving mathematics with a rather marginal status within general education for all in secondary schools. For most of the 19th century, there was no systematic teaching
Classes lmentaires
calcul calcul
Classes de grammaire
Optional:
Troisime
Confrences
Seconde
math-
Rhtorique
matiques
Mathmatiques lmentaires
Philosophie
Cours de mathmatiques accessoires
Mathmatiques spciales et Physique
FIGURE 1 Structure of mathematics instruction in French secondary schools in 1845 [Hahn 1846]
THE FIRST INTERNATIONAL COMMISSION
51
of mathematics for developing mathematical knowledge throughout the school years rather, it was reduced to a few basic notions in the rst years and to some concentrated instruction in the nal years. As Figure 1 shows, mathematics education in French secondary schools had practically returned as far as mathematics teaching for all was concerned to the Jesuit system where mathematics was not taught before the very end of the coll` ges : in the philosophy class. For a selection of students who e wanted more instruction there were optional lessons in the three senior classes. And for those who aspired to stand as candidates for the Ecole polytechnique there were special classes during the last two years. During the Third Republic, mathematics became more permanently established as a teaching subject for all yet still at the bare minimum for most classes, gaining importance only in the nal classes as the following Table shows for 1885 :
1885 Sixi` me e Cinqui` me e Quatri` me e Troisi` me e Seconde Rh torique e Philosophie 1 1 1 2 2 2 4
Number of weekly hours for mathematics instruction in the classes of French secondary schools [Belhoste 1995, 500 ff.]
Actually, the lack of continuity, as well as the marginal level of instruction, in mathematics and in the sciences, had become a matter of concern of educational policy since 1847. To supplement these deciencies in the classical lyc es, an enseignement secondaire sp cial was established, although the e e course was entirely concerned with practice and applications and hence of a lower status. The experiment, initiated in 1852, of a bifurcation between the humanities and the sciences within the lyc es had to be abandoned a few years e later [Hulin-Jung 1989]. The originally vocationally oriented enseignement sp cial then rose in status and achieved in 1891 (now renamed enseignement e secondaire moderne) a status equivalent to that of classical studies [Belhoste
52
G. SCHUBRING
1995, 226 and passim]. Eventually, in 1902, both branches merged to form just one school, but one which was internally differentiated into two, and then four, sections :
1902 Premier Cycle Section A Sixi` me e Cinqui` me e Quatri` me e Troisi` me e 2 2 1 + 1 opt. 2 + 1 opt. Section B 3 3 4 3
Second Cycle Sect. A et B Seconde Premi` re e Philosophie 1 1 2 optional Seconde Premi` re e Math matiques e Sect. C et D 5 5 8
Number of weekly hours for mathematics instruction in the classes of French secondary schools [Belhoste 1995, 577 ff.]
The reform of 1902 is famous in the history of French mathematics education for having established a decisive improvement of the status of mathematics as a teaching subject and for having introduced the notion of function as a key concept. As the Table above shows, this improvement of status affected only the science sections C and D, which constituted separate classes for the last three years. For these nal years, the lettres classes suffered strong reductions in the teaching of mathematics. And a closer scrutiny of the programmes shows that the concept of function was not introduced explicitly in the mathematics syllabus for the science sections, but only in the physics syllabus of the Seconde and even there only in the observations intended for teachers 1 ) :
1
) Clearly, in the programme for the nal class (Math matiques), the term function was e
used.
53
Conseils g n raux. Le professeur [] utilisera fr quemment les repr sene e e e e tations graphiques, non seulement pour mieux montrer aux el` ves lallure des ph nom` nes, mais pour faire p n trer dans leur esprit les id es si importantes e e e e e de fonction et de continuit . 2 ) [Belhoste 1995, 600] e
Curiously enough, the only explicit introduction of the function concept in a mathematics syllabus occurs in the programme for the nal class (Philosophie) of the lettres sections, where mathematics appeared only as an optional subject : notion de fonction, repr sentation graphique de fonctions e tr` s simples 3 ) [Belhoste 1995, 595]. e A decisive change, which reinforced mathematics as a permanent discipline, took place only after 1902. But mathematics still suffered from a split between general education and a better status for those students who aimed at later specialisation in mathematics. The scientic unity is said to have been achieved in 1925 by establishing a tronc commun for the humanities and the science sections, that means common teaching of both strands during the premier cycle. In contrast to the strictly centralised French state, GERMANY was divided up into a host of separate territories : 39 since 1815, and still 25 after the establishment of the German Reich in 1871, with the exclusion of Austria. Actually, this German Empire was formally a mere confederation of independent states, and education was a major area where this independence was emphasised. Consequently, status, extent and organisation of mathematics instruction continued to differ considerably between these states. One structural pattern common to all these federal states, however, was that there were two or even three parallel types of secondary schools : varying between one type featuring the values of classical Antiquity and another, decidedly modern type. Evidently, the status of mathematics depended entirely on the degree of modernity represented by the respective school type. In Prussia, the dominating state of the Federation, mathematics enjoyed a relatively good position in the classical school type, with 3 to 4 hours per week, extending to 5 to 6 hours in the genuinely modern type of school, but mathematics had a rather restricted role in the other German states. In ITALY, which had become a unied state since 1861, mathematics suffered a decline unique in all of Europe : in the only and exclusively
2 ) General advice. The teacher [] shall make frequent use of graphical representations, not only in order to let the pupils get a better understanding of the shape of the phenomena, but so as to inculcate them with the utterly important notions of function and continuity. 3
) the concept of function, graphical representation of very simple functions.
54
G. SCHUBRING
classical type of secondary school of that time, mathematics was conceived as a classical subject, as well, with Euclid as the ofcial text-book, together with the consequent frightening requirements of rigour. This teaching methodology resulted in ever repeated student failures, in particular in the nal examination. The solution, eventually adopted in the 1880s, was to abolish the character of mathematics as a major discipline and to reduce it to the status of a minor discipline and so remove it from the nal examinations. ENGLAND is somewhat analogous to the Italian case. It was not until the 19th century that some secondary schools, in the form of the famous public schools actually independent, private establishments , rst began to constitute a separate secondary school system, thus differentiating schools from the formerly all-embracing university colleges and arts faculties. Mathematics instruction there was dominated by closely following Euclid as a text-book and as a methodology for formalist elementary geometry. These few examples might be sufcient to show that around 1900 the national situations regarding mathematics teaching were so profoundly different that the absence of international communication can be easily understood. From the very moment, however, that the newly launched journal LEnseignement Math matique began to publish reports describing experiences with e mathematics teaching in some countries, the demand for international communication began to rise. This demand was transformed into a concrete proposal by David Eugene Smith in 1905. Smith, professor for mathematics teaching in New York, was the rst to propose an institutionalisation, not only of communication but, much more, of international cooperation. His proposal was :
La meilleure mani` re de renforcer lorganisation de lenseignement des e math matiques pures, serait de cr er une commission qui serait nomm e par e e e un Congr` s international et qui etudierait le probl` me dans son ensemble. 4 ) e e [Smith 1905, 469]
Thus, the journal LEnseignement Math matique has not only the seminal e merit of having established international communication on mathematics teaching, it also encouraged taking the next step, i.e. establishing cooperation. Due to this excellent groundwork, Smiths proposal was adopted when he renewed it in 1908 at the next International Congress of Mathematicians. Following Smiths proposal, meanwhile complemented by some issues concerning school
4 ) The best way to reinforce the organisation of the teaching of pure mathematics, would be the establishement of a committee appointed by an international Congress and which would study the problem in its entirety.
55
curricula which such a committee might investigate, the Fourth International Congress of Mathematicians in Rome voted to establish an international committee on mathematics instruction, with a mandate until the next Congress to take place four years later. A skeleton team of three was nominated : Felix Klein from Germany, Henri Fehr from Geneva and George Grenhill from Great Britain. This team, named Comit Central, a common term at the e time and without the later political connotations, elected Klein president and Fehr secretary-general to the committee. This did not only acknowledge the role of LEnseignement Math matique in launching this cooperative endeavour, e the journal also had a decisive function in realising the tasks of cooperation towards an international effort in favour of mathematics education. In fact, the new committee, whose ofcial name was used either in its German version : IMUK (Internationale mathematische Unterrichtskommission), or in its French version : CIEM (Commission internationale de lenseignement math matique), e decided that LEnseignement Math matique was to become the ofcial journal e for the entire work of IMUK/CIEM. One of the rst issues facing the Comit Central was to decide which e countries should be invited to participate in the work of IMUK. This question had in principle already been answered by the number of participating members of the ICM with voting rights. Its application to the work of IMUK, however, strikes us today as rather bizarre and the Comit Central itself did not appear e to have been satised with this crude denition of cooperation. The intention was to try to tie a given countrys level of activity in mathematics education to its participation in international mathematical communication : this level being determined by the number of participants at the four international mathematical congresses between 1897 and 1908. Countries having sent at least two mathematicians to at least two of the four congresses (Group 1) were invited to constitute a national subcommittee of IMUK and to be represented in the IMUK-body by one voting member ; for those whose number of participants had been ten or more, the number of voting members was raised to two or even three (Group 2). It is not clear to me how this arithmetic of proportion was determined. One possible reason is that this proportionality might have served to legitimise the IMUK endeavour to the respective government : it would be unimaginable today for national subcommittees to be semi-ofcial bodies funded by their respective countries where members had to be approved by the government of their country. For a reason yet unknown to me, the British
56
VOTING GROUP 2
G. SCHUBRING
MEMBERS OF THE
IMUK
GROUP 1 Belgium Denmark Greece Netherlands Norway Portugal Rumania Spain Sweden
Austria British Isles France Germany Hungary Italy Russia Switzerland USA
government even went so far as to exercise a certain supervisory role in approving the respective delegates. These criteria yielded a sample of eighteen countries, with nine countries in each of the two groups. As the Comit Central realized that the practice of e mathematical activity as decided by the ICM would not embrace all countries engaged in mathematics education, it was decided to invite the following countries to participate with non-voting members. The list (in italics are those countries which demonstrated at least temporarily a certain degree of activity) shows that by now all continents were involved. As one of Kleins assistants in the IMUK work reported, the idea had been to have the civilized countries represented in the IMUK body [Schimmack 1911, 2].
NON-VOTING, Argentina Australia Brazil British India Bulgaria
INVITED COUNTRIES
Canada Cape Colony Chile China Egypt
Japan Mexico Peru Serbia Turkey
Besides Henri Fehr, a considerable merit in constituting this rst international network is due to the efforts of D. E. Smith. Through advising F. Klein, he succeeded in establishing a great number of the personal contacts necessary for setting up the various national subcommittees. We can have some understanding of the tasks undertaken by IMUK, and make some assessment of them, by considering the main players, namely
57
the members of IMUK, and what constituted their relation to mathematics education. As can be seen from the list in Appendix 1 showing the members (delegates) in 1914, the great majority of members were university professors of mathematics. Their relation to mathematics teaching was constituted by the fact that at least a considerable number of their students were intending to become secondary school teachers. Probably the only professional among them was David Eugene Smith who, from the beginning of his career, had worked in normal schools, i.e. at teacher training institutions in the USA 5 ). His productivity was connected with his professorship at the Teachers College, a part of Columbia University in New York. His exclusive task there was to train mathematics teachers. He succeeded in rmly establishing mathematics education, by his own reections on teaching methodology, by publishing school textbooks, and by extensive historical research. The fact that the overwhelming majority of IMUK members came from higher education constituted at the same time a strength and a weakness of IMUKs work. It was strong in that these members were rmly rooted within the mathematical community. The rst IMUK thus worked on educational problems as an integral part of mathematical issues there were no conicts, no divergences between mathematicians and educators. On the other hand, this enracinement within the mathematical community was a weakness of the rst IMUK, since work was not undertaken from a proper perspective of school mathematics, i.e. that of teaching mathematics in schools. This potential weakness did not become an obstacle or a problem, however : rstly, since proper communities of mathematics education had not yet emerged, and secondly since Felix Klein had successfully avoided the one-sidedness which seemed to be an inevitable outcome of the domination of IMUK by higher education and by mathematicians concerns. In fact, the consequence of that domination was that the ICM, when creating IMUK, had commissioned it to deal with mathematics instruction at secondary schools only. This bias in favour of secondary schools was a result of the fact that university mathematicians considered this sector as the only eld where they were possibly educationally competent. Felix Klein, and the Comit Central with him, modied this commission and undertook to extend e their task to all school levels : from primary schools up to universities and technical colleges, and even including vocational schools !
5 ) Cf. the thesis by Donoghue [1987] for an account of Smiths growing engagement in mathematics education.
58
G. SCHUBRING
Dans le texte m me de la r solution du Congr` s de Rome il nest question que de e e e lenseignement math matique dans les ecoles secondaires . Mais, etant donn e e ` que le but de ces ecoles et la dur e de leurs etudes est tr` s variable dun Etat a e e lautre, le Comit fera porter son travail sur lensemble du champ de linstruction e math matique, depuis la premi` re initiation jusqu` lenseignement sup rieur. e e a e ` Il ne se bornera pas aux etablissements dinstruction g n rale conduisant a e e lUniversit , mais il etudiera aussi lenseignement math matique dans les ecoles e e techniques ou professionnelles. [] Il sagit donc dune etude densemble de lenseignement math matique e ` dans les diff rents types d coles et a ses divers degr s, [] 6 ) [CIEM 1908, e e e 451452]
The work of IMUK and of its national subcommittees on mathematics instruction at all these school levels was really impressive. The ofcial list of publications established in 1920, upon the dissolution 7 ) of IMUK, documents 294 contributions published in 17 countries. Evidently, the quality of these papers varies signicantly. In some countries, the reports were the result of truly collective and intensive work while, for other countries, the reports were prepared by individuals. The German reports were generally noted for being the best organised. Communication and cooperation had not been an aim in itself for the Comit ; they were understood rather as a process with a direction, namely e initiating reforms. The pivotal point for launching the rst international reform movement in mathematics education was, in fact, the decision of the Comit e Central to complement the (more or less descriptive) national reports submitted by the national sub-committees with international comparative reports on a few key topics representing the major reform concerns. These topics, perhaps unsurprisingly, reected the fact that the viewpoints of higher education were indeed dominant in the rst IMUK. Of the eight topics for written reports, only three (1, 2, 5) corresponded to the level ofcially designated as IMUK responsibility, i.e. the level of secondary schools, whereas the majority of ve topics concerned either the transition from secondary to higher education or
6 ) In the text adopted by the Rome Congress, only mathematical teaching in secondary schools is mentioned. But, given that the aims of these schools and the duration of their studies are quite different from state to state, the Committee will extend its work to the entire eld of mathematical instruction, from the rst initiation to superior education. It will not limit itself to institutions of general education leading to University, but it will also study mathematical teaching in technical or vocational schools. [] We will therefore undertake a global study of mathematical teaching in the various types of schools and at its various levels, [] 7 ) For more information on the different periods of existence of IMUK, see the additonal comments made by the author during the general discussion on this theme, as reported in this volume (p. 91).
59
related exclusively to higher education. The titles of the reports 8 ) were :

1. La rigueur dans lenseignement math matique dans les ecoles moyennes e (Milan, 1911). 2. La fusion des diff rentes branches math matiques dans lenseignement dans e e les ecoles moyennes (Milan, 1911). 3. Lenseignement math matique th orique et pratique destin aux etudiants en e e e sciences physiques et naturelles (Milan, 1911). ` 4. La pr paration math matique des physiciens a luniversit (Cambridge, 1912). e e e 5. Lintuition et lexp rience dans lenseignement math matique dans les ecoles e e moyennes (Cambridge, 1912). 6. Les r sultats obtenus dans lintroduction du calcul diff rentiel et int gral e e e dans les classes sup rieures des etablissements secondaires (Paris, 1914). e 7. La pr paration math matique des ing nieurs dans les diff rents pays (Paris, e e e e 1914). 8. La formation des enseignants (professeurs) des math matiques pour les e etablissements secondaires. 9 )
On the other hand, it is quite clear that these topics, which helped to initiate reform movements in many of the participating countries, constituted essential concerns for the relation between mathematics education and mathematics, and they have remained pivotal points for this relation even today. I shall give an overview of the major contents of these thematically comparative reports : 1. RIGOUR The rapporteur for this topic was Guido Castelnuovo from Italy. He conned himself to geometry. While efforts in the other countries were towards reducing rigour in teaching geometry, while introducing empirical methods to encourage the role of intuition, such reforms were refuted in principle in Italy [Castelnuovo 1911]. The majority of Italian mathematicians even opted, and acted, for reinforcing rigour in geometry. A presentation of geometry in a more and more axiomatic form over the school years is characteristic of this Italian approach.
8 ) In brackets we mention the year and the place of their presentation either to a general IMUK meeting or to an IMC. See also Appendix 3 in [Furinghetti 2003]. 9 ) 1. Rigour in mathematics teaching in secondary schools. 2. The fusion of the various mathematical branches in secondary school teaching. 3. Pure and applied mathematics teaching for students in physics and the natural sciences. 4. The mathematical preparation of physicists at university. 5. Intuition and experimental evidence in mathematics teaching in secondary schools. 6. Results obtained from introducing calculus in the upper grades of secondary institutions. 7. The mathematical preparation of engineers in the various countries. 8. The training of mathematics teachers for secondary institutions. The last study, decided in 1914 for the next session, was delayed due to World War I and was presented only at the 1932 Congress.
60 2. FUSION
G. SCHUBRING
The report on the second topic concentrated on the controversy whether it was preferable or not for geometry teaching to integrate planimetry and stereometry. Other possible concrete realisations of fusion, say, of integrating algebra and geometry, were not discussed. Although the rapporteur was a Frenchman, Bioche, the discussion was again dominated by the Italians. The controversy whether planimetry and stereometry should be taught in an integrated manner, or not, had been dividing the Italian mathematical community for decades [Bioche 1911]. The pivotal point of the controversy was the concept of the purity of the mathematical method. Since the Italian syllabus of 1906, which admitted a plurality of teaching methods, had been replaced in 1910 by an anti-fusionist one, the traditional purists in Italy had eventually won. 3. PURE
AND APPLIED MATHEMATICS FOR STUDENTS OF THE SCIENCES
The rapporteur was this time a German, Heinrich Timerding. He presented his report along the lines of the German problematic : the hotly debated point being whether the mathematical formation necessary to study the natural sciences should be specially adapted to the future profession of the students, or whether a general education in mathematics as a common foundation for the study of the sciences was preferable. Timerding criticised the too narrow specialisation prevailing in Germany, emphasising the more general character of mathematical education in France and Italy [Timerding 1911]. Since the two topical reports to the Cambridge Congress of 1912, where the mandate of IMUK was prolonged for another four years, essentially discuss variations of the rst three topics, I will now concentrate on the last two, presented at Paris in 1914. It is quite remarkable that the culminating point of the IMUK work was reached in 1914 at its Paris session where the two topics presented corresponded exactly to the two cornerstones of Kleins German reform programme. In Paris, the subject that attracted the most attention and participation was the evaluation of the introduction of calculus to secondary schools. The topic was hotly debated, and the report on it was the most voluminous of all the international IMUK reports [Beke 1914]. This was also the topic that Klein had prepared more carefully than any other. He not only helped design the international questionnaire that dealt with this matter, but he also chose
61
its coordinator and reporter, Emanuel Beke, a Hungarian scholar and former student of Klein, and one of the most fervent adherents of Kleins programme. Beke was responsible for transmitting the reform ideas directly to Hungary and he was highly successful in initiating an analogous movement there. For the second topic of the Paris session, the mathematical preparation of engineers, Klein also chose a trustworthy person as chairman, namely Paul St ckel, a close co-worker of Kleins in matters of history and of applied a mathematics [St ckel 1914]. Klein, who did not participate in the Paris session a himself, was immediately informed about the course of the debates by letters from his assistants and co-workers. Concerning the mathematical training of engineers, Lietzmann was able to report, much pleased, that the engineers want this was the general opinion to get their mathematics from the mathematician, not from the engineer [Schubring 1989, 190]. St ckel also reported that he was quite a satised with the international response, particularly from the French engineers, the overwhelming majority of whom had stressed the necessity of a culture g n rale for engineers [ibid.]. e e Although Klein and his co-workers had expected more palpable results with regard to Calculus, the impact of his main reform agenda namely to introduce the concept of functional thinking as a basic notion pervading the entire mathematical curriculum proved to be enormously successful on the international level. Sooner or later, the syllabus became modernised almost everywhere, supplanting the traditional restriction to the static ideas of elementary geometry which excluded any knowledge of variables or any kind of modelling of physical processes. When Klein had begun to understand that not only had teacher education to be improved, but also that the school curriculum needed to be profoundly modernised, he criticised the deep chasm that existed between research mathematics and school mathematics. His intention in launching the reform movement had been to bridge that gulf, but in a manner which nonetheless acknowledged the structural difference between the domains of school and research. Klein used a tting metaphor to describe this necessary difference, adopting the term HYSTERESIS from physics, saying that there must necessarily be a certain hysteresis between the evolution of mathematical knowledge and the contents of school mathematics. This is needed to restructure the building of mathematics by integrating the new knowledge and achieving a new elementarisation of it [Klein 1933, 221]. Kleins view was that a hysteresis of about 30 years was a reasonable time span, and he criticised the fact that at the beginning of the 20th century the time lag amounted to rather more than
62
G. SCHUBRING
100 years, with the result that there was no more elasticity (to follow the physics metaphor) and that the relation between mathematics and the school curriculum had been completely severed. Upon reecting on the evolution of school curricula in mathematics across the 20th century, we observe that, as an effect of the work and initiatives of LEnseignement Math matique and of IMUK, such an immense hysteresis did e not occur again and that, conversely, mathematics education began to follow the great achievements of mathematics quite closely, the time lag of 30 years suggested by Klein becoming even shorter. School curricula proved able to develop in a highly dynamical manner. I should like to recall that applied mathematics was almost non-existent in Kleins time. It is due to efforts initiated in Gottingen, and rst realised by Carl Runge, that numerical mathematics began to evolve. Probability theory itself emerged as a distinct branch only since the 1920s. Today, however, probability theory and stochastics are not only a standard branch of secondary school mathematics, but basic probabilistic notions can even be found in the content of primary school mathematics courses. I should also like to mention the importance of the introduction of settheoretical notions to school mathematics. While so-called modern mathematics is seen only negatively today, we should note that the new syllabuses based on the notions of mathematical structures brought about the modernisation of the syllabuses for primary schools for the rst time. As other examples of quite recent developments in mathematical research which quickly found their way into school, I just mention non-standard analysis and fractals. Perhaps it would be well not to mention the immediate introduction of information technology to schools, since mathematics itself has problems defending its own position versus computer science. Thus an evaluation from an historical perspective allows us to see that the objective basis for the convergence between mathematics education and mathematics itself has been decisively improved, and that the two disciplines can contribute to fruitful cooperation from very advanced levels of scientic development. On the other hand, it seems to be a problem for many mathematicians to understand that learning processes are not determined by content alone, and that research into learning processes requires very broad conceptual and empirical work, at least where the particularly resistant subject of mathematics is concerned. The enormous progress in research on pedagogical, psychological and epistemological foundations of the learning process in mathematics is responsible for the emergence of the didactics of mathematics as a scientic
63
discipline in its own right. For the moment, it has to be admitted, this development carries with it the danger of an estrangement (Entfremdung) between the mathematics community and mathematics education. Not only is this danger visible in many countries where mathematics educators have only a precarious foothold within the mathematics departments, but it has even become acute where mathematics education is well established. As an example we can cite the so-called mathematics war in California in recent years, where mathematicians have taken up arms against mathematics educators 10 ). Although it would seem that these mathematicians were not aware of their being manipulated by politicians who tried to maintain elitist concepts of education, I nd it deplorable how strongly they favoured highly conservative and out-dated concepts of drill and practice and learning by rote. Given the objective basis for a convergence between mathematics and mathematics education, and the two disciplines mutual interest in a qualitative improvement of school mathematics and of learning mathematics, I hope that the evaluation of these One Hundred Years will contribute towards promoting and improving these relations.
10 ) See the controversial statements by Henry L. Alder, Jerry P. Becker and Bill Jacob in : ICMI Bulletin 44 (1998), 1625 and 45 (1998), 916. The conicts are not conned to mathematics but concern other school subjects, too (see [Stotsky 2000]).
64
G. SCHUBRING
APPENDIX 1 MEMBERS
OF
IMUK / CIEM
IN
1914
Australia. Horatio Scott Carslaw (univ.) Austria : Emanuel Czuber (univ) ; Wilhelm Wirtinger (univ.) ; Richard Suppantschitsch (univ.) Belgium : J. Neuberg (univ.) Brazil : Raja Gabaglia (sec. school) British Isles : Sir George Greenhill (Artillery college) ; E.-W. Hobson (sec. school) ; Charles Godfrey (Naval College) Bulgaria : A. V. Sourek Cape Colony : S. S. Hough (Observatory) Denmark : Poul Heegard (univ.) Egypt : F. Boulad France : Jacques Hadamard (Polytechn., Coll` ge de France) ; Maurice dOcagne e (Polytechn.) ; Charles Bioche (sec. school) Germany : Felix Klein (univ.) ; Paul Staeckel (univ.) ; Albrecht Thaer (sec. school) Greece : Cyp. St phanos (univ.) e Italy : Guido Castelnuovo (univ.) ; Fed. Enriques (univ.) ; Gaetano Scorza (univ.) Japan : Rikitaro Fujisawa (univ.) Mexico : Valentin Gama (Observatory) Netherlands : J. Cardinaal (polytechn.) Norway : Olaf Alfsen (sec. school) Portugal : Gomes Teixeira (Polytechn.) Rumania : G. Tzitzeica (univ.) Russia : Nikolaj v. Sonin (univ.) ; Boris Kojalovic (polytechn.) ; Constantin Poss (univ.) e Serbia : Michel Petrovitch (univ.) Spain : Luis-Octavio de Toledo (univ.) Sweden : Edvard G ransson (sec. school) o Switzerland : Henri Fehr (univ.) ; Carl Friedrich Geiser (Polytechn.) ; J.-H. Graf (univ.) USA : David Eugene Smith (Teachers College) ; William Osgood (univ.) : John Wesley Young (univ.)
65
REFERENCES BEKE, E. Les r sultats obtenus dans lintroduction du calcul diff rentiel et int gral e e e dans les classes sup rieures des etablissements secondaires. Rapport g n ral. e e e LEnseign. Math. 16 (1914), 245284. BELHOSTE, B. (ed.). Les sciences dans lenseignement secondaire francais : textes ofciels, r unis et pr s. par Bruno Belhoste. Tome 1 : 17891914. Inst. Nat. de e e Recherche P dagogique, Paris, 1995. e BIOCHE, CH. Rapport de la Sous-commission : La question de la fusion des diff rentes e branches math matiques dans lenseignement moyen. LEnseign. Math. 13 (1911), e 468471. CASTELNUOVO, G. Rapport de la Sous-commission : La rigueur dans lenseignement math matique dans les ecoles moyennes. LEnseign. Math. 13 (1911), 461468. e CIEM. Rapport pr liminaire sur lorganisation de la commission et le plan g n ral de e e e ses travaux. LEnseign. Math. 10 (1908), 445458. DONOGHUE, E. F. The origins of a professional mathematics education program at Teachers College. Ed. D. Thesis, Teachers College, Columbia University, New York, 1987. FURINGHETTI, F. [2003] Mathematical instruction in an international perspective : the contribution of the journal LEnseignement Math matique. This volume, 1946. e HAHN, L. Das Unterrichtswesen in Frankreich : mit einer Geschichte der Pariser Universitt. Gosohorsky, Breslau, 1846. HULIN-JUNG, N. Lorganisation de lenseignement des sciences : la voie ouverte par le Second Empire. Ed. du Comit des Travaux historiques et scientiques, Paris, e 1989. KLEIN, F. Elementarmathematik vom hoheren Standpunkte aus. Band 1 : Arithmetik ; Algebra ; Analysis. 4. Au. Springer, Berlin, 1933. LEHTO, O. Mathematics Without Borders : A History of the International Mathematical Union. Springer, New York, 1998. SCHIMMACK, R. Die Entwicklung der mathematischen Unterrichtsreform in Deutschland. Teubner, Leipzig, 1911. SCHUBRING, G. Pure and applied mathematics in divergent institutional settings in Germany : the role and impact of Felix Klein. In : The History of Modern Mathematics. Vol. II : Institutions and Applications, eds. D. Rowe, J. McCleary, 171220. Academic Press, Boston, 1989. ` SMITH, D. E. Opinion de David-Eugene Smith sur les r formes a accomplir dans e lenseignement des math matiques. LEnseign. Math. 7 (1905), 469471. e STACKEL, P. La pr paration math matique des ing nieurs dans les diff rents pays. e e e e Rapport g n ral. LEnseign. Math. 16 (1914), 307328. e e STOTSKY, S. (ed.). Whats at Stake in the K-12 Standard Wars. Peter Lang, New York, 2000. TIMERDING, H. Rapport de la Sous-commission : Lenseignement math matique e th orique et pratique destin aux etudiants en sciences physiques et naturelles. e e LEnseign. Math. 13 (1911), 481496.
JOURNALS OF MATHEMATICS EDUCATION, 19002000 ` Les revues denseignement des math matiques, de 1900 a 2000 e par Gila HANNA
On examine dabord l volution du nombre des revues denseignement des e math matiques au si` cle dernier. Puis on discute les caract ristiques de trois importantes e e e revues internationales qui sont enti` rement d di es a la recherche en didactique des e e e ` math matiques. e Dans la premi` re partie on pr sente une recherche sur les revues, effectu e e e e au moyen des sites web de Ulrichs International Periodicals Directory [UIPD] et de Zentralblatt f r Didaktik der Mathematik [ZDM] (connu aussi sous le nom u dInternational Reviews on Mathematical Education). En combinant et comparant les r sultats de la recherche sur les deux sites on voit que le nombre total de revues e ` sur lenseignement math matique est pass de six au commencement du si` cle a e e e ` 314 en 1999. Nous analysons la distribution statistique des revues selon les pays o u e elles ont et publi es et la d cennie de leur premi` re publication. Le caract` re des e e e e revues est etudi a partir des d clarations editoriales : on a trouv plus de similarit s e ` e e e ` que de diff rences. Selon ces d clarations la majorit des revues se consacrent a la e e e ` compr hension et a lam lioration de lenseignement des math matiques, elles publient e e e des articles qui peuvent int resser les enseignants et les chercheurs, et elles sont e ` ouvertes en particulier a des contributions qui discutent des nouvelles orientations dans la recherche en didactique des math matiques ou des potentialit s des nouvelles e e technologies pour am liorer lenseignement des math matiques. e e Dans la seconde partie de cette contribution, on analyse plus en d tail trois e importantes revues : Educational Studies in Mathematics (ESM), Journal for Research in Mathematics Education (JRME) et For the Learning of Mathematics (FLM). Les deux premi` res revues sont compar es a partir de leurs prols initiaux, leur d claration de e e ` e politique editoriale et les sujets trait s dans les articles. ESM et JRME commenc` rent e e dans la m me p riode (en 1968 et 1970 respectivement), mais elles ont pris des e e directions diff rentes dans les deux premi` res d cennies de leur existence. ESM publiait e e e des discussions et des essais, tandis que JRME publiait des articles bas s sur la recherche e
68
G. HANNA
` exp rimentale et sur lanalyse quantitative de donn es (op rant surtout a partir de tests e e e ` a choix multiples). Cette situation changea par la suite dune facon signicative, bien que gradu ellement. A la n des ann es 1970, ESM commenca a publier une grande vari t e ` ee darticles, y compris certains qui traitaient de recherche quantitative bas e sur des e donn es statistiques. Aussi le sujet des articles changea-t-il consid rablement. M me e e e si la plupart des articles etaient encore des essais conceptuels sur des aspects de la ` didactique des math matiques, la revue se mit a publier des rapports de recherche sur e les probl` mes de la classe comme le style dapprentissage, le travail en petits groupes, e la diff rence de genre, linteraction dans la classe, les attitudes des etudiants et les e interpr tations fausses. A partir de la n des ann es 1980, JRME diminua le nombre e e darticles bas s sur les projets exp rimentaux, linf rence statistique et autres m thodes e e e e ` quantitatives. Dans le m me temps on assista a un accroissement dans la proportion e darticles bas s totalement ou en partie sur des m thodes qualitatives. Lorientation e e vers une convergence de contenus dans les deux revues a continu au cours de la e troisi` me d cennie. e e FLM est la plus jeune des trois revues, puisquelle est n e en 1980. La diff rence e e entre FLM et les deux autres revues r side dans le fait quelle publie souvent des e articles centr s sur des disciplines qui, dans la tradition, ne sont pas associ es a e e ` lenseignement math matique, comme la sociologie, la linguistique, lhistoire, lart, la e philosophie et lethnomath matique. e
JOURNALS OF MATHEMATICS EDUCATION, 19002000 by Gila HANNA
Today, at the beginning of the 21st century, mathematics education can boast of hundreds of journals, published all over the globe and in many languages. Our historical data reveals very few journals of mathematics education on the threshold of the 20th century, when the French-language journal LEnseignement Math matique was rst published in Switzerland. This e certainly reects the status of mathematics education at the time, when it was not yet recognized as a scholarly discipline. As mathematics education slowly acquired such a status, however, rst in Europe and then in North America and the rest of the world, many new journals of mathematics education began to appear, and the trickle of new journals at the turn of the 20th century turned into a ood by mid century and still continues. This growth in the number of journals over the last century reects a signicant need created by the growth in the eld of mathematics education itself. In this eld, as in many others, scholarly journals are the primary vehicle for the dissemination of new research ndings and the primary forum for the exchange of views and methods by its practitioners. In addition, they play a crucial role in helping the professional community monitor the quality of its research and aggregate the disparate pieces of new knowledge. It is difcult to arrive at a precise count of journals of mathematics education, because the relevant data-bases do not always use clear descriptors and unambiguous keywords. In addition, the data itself can only reect the necessarily limited manner in which journals choose to represent themselves in their titles and keywords. Several educational journals, for example, do not have the word mathematics in their title or list of descriptors, but nevertheless do publish many articles on mathematics education (and even occasional special issues entirely devoted to it). On the other hand, of course, many journals that do have the word mathematics in their title never publish articles on mathematics education.
70
G. HANNA
ULRICHS INTERNATIONAL PERIODICALS DIRECTORY A search 1 ) of the electronic, on-line version of Ulrichs International Periodicals Directory [UIPD 2000] (which appears to provide data for up to 1999) yielded a list of 209 journals, published in 28 countries. Conrmation that this count is more or less correct came from a search of the list for 1999 of some 400 serials covered by Zentralblatt fur Didaktik der Mathematik : International Reviews on Mathematical Education [ZDM 2000], in the areas of mathematics, mathematics education, and general education. This search yielded about 200 journals of mathematics education. It should be pointed out that Ulrichs directory is incomplete. It does not list, for example, such well-known journals of mathematics education as For the Learning of Mathematics (Canada) and La Matematica e la sua Didattica (Italy), and it reports nothing for some countries for which it is unreasonable to assume that there are no journals of mathematics education in any local language. Its data-base contains important details on each journal reported, however, such as rst year of publication, frequency of appearance, publisher, whether refereed or not, cessation of publication, and a short description of the purpose of the journal. Accordingly, the 38th edition of Ulrichs directory is the most comprehensive source available, and nearly all the information presented below is based upon it. Refereed journals Of the 209 journals on the list of Ulrichs directory, 61 were shown as refereed. This means only that at least 61 are refereed, as this label was obviously attached only to journals that had made a point of conrming that they were refereed. Mathematical Thinking and Learning : an international journal and the American Mathematical Monthly, for example, both of which are known to be refereed journals, were not recorded as such on this list. Status The status of the 209 journals listed in [UIPD 2000] was as follows : Active : 169 Ceased : 24 Unveried : 16
The 24 ceased journals were distributed as follows : Australia (1), Canada (1), France (1), Germany (5), India (1), Papua New Guinea (2 : actually it was
1
) using the description mathematics
AND
(education
OR
learning
OR
teaching)
JOURNALS OF MATHEMATICS EDUCATION, 19002000
71
TABLE 1 Distribution of mathematics education journals by country (according to Ulrichs directory) Australia Botswana Canada China Czech Republic Finland France Germany India Indonesia 22 1 6 9 3 1 10 17 4 1 Italy Japan Korea Malaysia Netherlands Nigeria Papua New Guinea Poland Russia Saudi Arabia 4 19 1 5 8 1 2 6 2 2 Spain Sri Lanka Sweden Switzerland Taiwan Tanzania United Kingdom United States Total Active 1 1 1 4 1 1 21 55 209 169
only one, appearing with different titles), Poland (1), United Kingdom (3) and United States (9). For ten of these, there was no indication of a terminal date. The other fourteen had been published for as long as 98 years (one journal, in France) and as little as four years (two journals, in the USA). The median life of the 14 journals was 11 years, with a minimum of 4 years and a maximum of 27 (excluding the journal that lasted 98 years). Distribution of journals by decade of rst publication For 33 of the 209 journals in Ulrichs directory, there was no indication of the year in which they were rst published. Therefore Figure 1, which shows the distribution of the journals by decade, covers only those 176 journals for which the year of rst publication was given. As mentioned, Ulrichs directory is clearly incomplete. Schubring [1980] lists 253 journals, and Appendix 1 of [Furinghetti 2003] indicates that over 70 journals were listed in the Bulletin Bibliographique of LEnseignement Math matique as appearing in the years 18991901. The names of these e journals would seem to indicate that most of them were devoted primarily to pure and applied mathematics. It would be fair to assume, however, that a number of them did publish papers on mathematics education, and even that some of them were dedicated to that topic, so that the numbers given in Figure 1 for these years, on the basis of Ulrichs directory, can only be an underestimation. For example, important journals such as Periodico di matematica (founded in 1886), Il Bollettino di Matematica (founded in 1902), and Bollettino della
72
G. HANNA
1 9 9 0 -1 9 9 9 1 9 8 0 -1 9 8 9 1 9 7 0 -1 9 7 9 1 9 6 0 -1 9 6 9 1 9 5 0 -1 9 5 9 1 9 4 0 -1 9 4 9 1 9 3 0 -1 9 3 9 1 9 2 0 -1 9 2 9 1 9 1 0 -1 9 1 9 1 9 0 0 -1 9 0 9 -1 8 9 9 0 1 2 6 10 20 30 2 4 4 25 25
35 34 38
40
FIGURE 1 Distribution of 176 journals by rst year of publication within each decade (according to Ulrichs directory)
Mathesis (founded in 1909), do not appear in Ulrichs directory even though they are written in Italian, one of the major languages used in mathematical journals and conferences of the period. Only six journals included in Ulrichs directory began publication prior to 1900. They are : 1) Statistics in Society (in English), 1838. It is published three times a year by the Royal Statistical Society, Blackwell Publishers Ltd., United Kingdom. This refereed journal carries papers on economic, social and governmental issues (historical, biographical, philosophical, demographical) and on medical statistics, but appears to carry few papers related to mathematics education. 2) American Mathematical Monthly (in English), 1894. It is published ten times a year by the Mathematical Association of America, and is currently edited by John H. Ewing. It publishes expository articles on all facets of mathematics, pure and applied, old and new, and on teaching approaches, with regular columns devoted to reviews and to both basic and complex problems.
73
3) LEducation Math matique (in French), 18981980 (ceased). It was e published by Librairie Vuibert, Paris, France. It is described as a journal that published articles on mathematics, education, teaching methods, and curriculum. 4) LEnseignement Math matique (in French), 1899. This is an academic, e scholarly journal currently published in Switzerland. 5) Revue des Math matiques de lEnseignement Sup rieur (in French), e e 1890. Before 1998 this publication was called Revue de Math matiques e Sp ciales. It is now published ten times a year. This publication carries e information on teaching at the undergraduate and rst graduate levels in France. 6) Mathematical Gazette (in English), 1894. This journal is currently published three times a year by the Mathematics Association, United Kingdom. It features articles about teaching and learning mathematics, with a focus on the 1520 age range, as well as expository articles on attractive areas of mathematics. In Figure 1, the [UIPD] data is presented by decade. In the rst decade, 19001909, two new journals appeared : Mathematics Teacher in 1901 and School Science and Mathematics in 1908. The rst, published by the National Council of Teachers of Mathematics (USA), contains papers on the improvement of mathematics instruction in junior and senior high schools, colleges, and teacher education colleges. The second, published by the School Science and Mathematics Association (USA), carries articles on assessment, curriculum, research, teacher education, learning theory, philosophy and history of science, non-traditional instruction, and the role and inuence of science and technology in society. Between 1910 and 1919 only one new journal appeared on the scene, according to this data-base : Japan Society of Mathematical Education Journal (in Japanese), 1919. It is published by the Japan Society of Mathematical Education and is described as being a mathematics education journal. In the three decades from 1920 through 1949 ten new journals appeared, two in the 1920s and four in each of the next two decades. In the 1920s the two new journals were Euclides (in Dutch), 1925, and Rozhledy Matematickofyzikalni (in Czech and Slovak), 1921. In the 30s the new additions were : Dimensio (in Finnish, with summaries in English, Finnish and Swedish), 1937, Mathematics and Physics / Tokyo Kyokai Daigaku Rigakubu Kiyoa (in multiple languages), 1930, Matematika
74
G. HANNA
v Skole (in Russian), 1934, and Mathematics Student (text in English) published by the Indian Mathematical Society, 1933. The 40s saw the publication of the Australian Mathematics Teacher (in English), 1945, the Hokkaido University of Education Journal. Section 2A : Mathematics, Physics, Chemistry, Engineering (text and summaries in both English and Japanese), 1949, Matematyka (text in Polish), 1948, and Archimede, 1949 (text in Italian, continuation of Il Bollettino di Matematica founded in 1902). Over the next ve decades, from 1950 to 2000, an average of 30 new journals of mathematics education arrived on the scene every ten years. In the 1950s the 25 new journals were distributed among six countries as follows : Japan (10), Germany (5), UK (5), Poland (2), USA (2), and Czechoslovakia (1). In contrast, the 25 new journals which appeared in the 1960s were distributed among as many as thirteen countries and included such prominent journals as Educational Studies in Mathematics and Zentralblatt fur Didaktik der Mathematik. The 1970s saw the publication of 38 new journals in seventeen countries ; the highest concentration was in the USA (12), Australia (7) and Germany (6). Among the new journals were the well-known journals Journal for Research in Mathematics Education and Journal fur Mathematikdidaktik. Of the additional 34 new journals that appeared in the 1980s, 15 were in the USA, with the other 19 spread over 11 countries. The Australian Mathematics Education Research Journal and the American Journal of Mathematical Behavior are among these new journals. Also noteworthy in the 1980s is the appearance of the word computers in the title of several of the new journals. Not surprisingly, the 1990s saw even more new journals with the word computer in the title, among them the International Journal of Computers for Mathematical Learning. There were also a few journals devoted specically to mathematics teacher education, such as the Journal of Mathematics Teacher Education. Ulrichs data-base does not contain information on journals that started appearing in the year 2000. For this reason it does not mention very recent additions such as the Greek journal Themes in Education or the Canadian Journal of Science, Mathematics and Technology Education, whose rst issues have just appeared. [UIPD 2000] does not mention electronic journals, but a Web search revealed that there are about a dozen new free electronic journals on mathematics education posted on the Web.
75
DATA-BASE
THE KARLSRUHE ZENTRUM FUR DIDAKTIK
DER
MATHEMATIK
The list of periodicals covered by the data-base maintained and edited by the Zentrum f r Didaktik der Mathematik at Karlsruhe University [ZDM] was u also consulted to complement the incomplete information provided by Ulrichs directory.
TABLE 2 Distribution of active mathematics education journals, in [ZDM] but not in Ulrichs, by country Argentina Australia Austria Belgium Brazil Canada Colombia Croatia Czech Republic Denmark Finland France Germany Greece 2 5 3 3 5 8 1 2 1 3 1 13 12 1 Hong Kong Hungary India Italy Japan Korea Malaysia Mexico Mozambique Netherlands New Zealand Norway Philippines Poland 1 2 3 5 3 1 1 1 1 3 1 2 1 3 Portugal Romania Russian Fed. Singapore Slovakia South Africa Spain Sweden Switzerland Ukraine United Kingdom United States Yugoslavia Total 2 2 1 3 1 2 4 2 6 1 12 20 1 145
Though [ZDM] provides up-to-date information on the presently active periodicals, it does not give details, such as rst year of publication or whether journals are refereed or not. The [ZDM] data-base lists 215 mathematics education journals, 70 of which were already part of [UIPD]. The additional 145 journals missing from Ulrichs were distributed among 41 countries as described in Table 2. The total number of active mathematics education journals in both databases is 314 (169 in Ulrichs and 145 others in [ZDM]). By way of comparison, there are about 675 active mathematics journals and about 35 of these are free electronic journals [Jackson 2000].
76
G. HANNA
EDITORIAL
STATEMENTS
Editorial statements describe the purpose and scope of each journal and thus might be expected to help in determining differences among journals. These statements are remarkably similar to each other, however. For example, the Journal for Research in Mathematics Education (JRME) in its rst issue in 1971 stated :
[This journal] issues the invitation to researchers to submit papers which represent a coordinated approach to research on a major problem in learning, methodology, role of materials, etc. [] It is anticipated that such papers could serve as the basis for an exchange among mathematics education researchers.
This is very close to the statement in Educational Studies in Mathematics (ESM), rst published in 1969 :
[ESM] presents new ideas and developments which are considered to be of major importance to those working in the eld of mathematical education. It seeks to reect both the variety of research concerns within this eld and the range of methods to study them.
Again, this is not much different from the editorial statement of For the Learning of Mathematics, rst published in 1980 :
The journal aims to stimulate reection [] to generate productive discussion [] to promote criticism and evaluation of ideas and procedures current in the eld []
This is also quite close to the editorial statement of the recent journal Mathematical Thinking and Learning, which rst appeared in 1999 and which invites high-quality articles
that make a signicant contribution to mathematical learning, reasoning, and instruction. Articles that are proactive and challenge the status quo are particularly sought. Authors who analyze future-oriented problems and offer clear, researchable questions for addressing such problems are likewise encouraged to make a contribution to the journal.
But there were journals that wanted to emphasize how different they were from existing journals. For example, the Journal of Mathematical Behavior (previously known as Journal of Childrens Mathematical Behavior) which was created in 1971, stated that :
[JMB] felt that the then-existing journals and common practices left a sizable gap which needed to be lled. It seemed that two distinct paradigms existed concerning research in mathematics education, but only one of these was represented in journals. The alternate paradigm, with which this Journal is concerned, is (at least roughly) dened by [the postulate that behavior is explained by information processing].
77
The International Journal of Computers for Mathematical Learning, rst published in 1996, somewhat deviated from the conventional statements by saying that the mathematics discussed is not limited by current boundaries of the discipline and stating :
[This journal] is dedicated to the understanding and enhancement of changes in the nature of worthwhile mathematical work that can be performed by learners, teachers and practitioners. It publishes contributions that open the way to new directions, particularly those which recognise the unique potential of new technologies for deepening our understanding. Contributions are both theoretical and empirical, practical as well as visionary.
CLOSER LOOK AT THREE PROMINENT JOURNALS
In the following I will discuss in some detail three journals that have gained international prominence and are arguably prestigious journals. Rather than being professional journals aimed at teachers, all three are research journals devoted entirely to current research in mathematics education : Educational Studies in Mathematics (ESM), the Journal for Research in Mathematics Education (JRME) and For the Learning of Mathematics (FLM). Educational Studies in Mathematics (ESM) BEGINNINGS. Educational Studies in Mathematics, published in the Netherlands by Kluwer Academic Publishers, was founded by its rst editor, Hans Freudenthal. The rst two issues appeared together in May 1968, and consisted of the proceedings of a colloquium on mathematics education held in Utrecht in August 1967 and entitled How to teach mathematics so as to be useful. The colloquium was an activity of ICMI. This combined issue consisted of 22 articles, of which 13 were in English, eight in French and one in German. Most of the authors were well-known mathematicians and mathematics educators, among them H. Freudenthal (The Netherlands), A. Krygovska (Poland), H. B. Grifths (UK), T. Fletcher (UK) and A. Revuz (France). ESM publishes papers in both English and French, though most appear in English. POLICY STATEMENT. Oddly, ESM did not publish a policy statement initially. It is not until the appearance of volume 10, number 1, in February
78
G. HANNA
1979, when Alan Bishop took over the editorship of the journal, that such a statement was made public :
The journal presents new ideas and developments which are considered to be of major importance to those working in mathematics education. It seeks to reect both the variety of research concerns within this eld and the range of methods used to study it. It deals with didactical, methodological and pedagogical subjects rather than with specic programmes and organisations for teaching mathematics, and it publishes high-level articles which are of more than local or national interest.
The wording of this statement was slightly changed in 1990 when the editorship was put in the hands of a team of three editors, with Willibald Doerer as editor-in-chief. Again the statement was slightly modied ; the sentence All papers are strictly refereed and the emphasis is on high-level articles which are of more than local or national interest was added in 1996 when Kenneth Ruthven assumed the responsibility of editor-in-chief. SUBJECTS OF ARTICLES. In its rst issues, ESM published articles that can best be described as reections on mathematics and its teaching, as well as articles on psychological and pedagogical aspects of teaching. It sought to establish strong links among scientic, educational and psychological research, mathematics, and classroom teaching. Most articles were of the essay form, presenting their arguments without relying on empirical educational data. There were almost no articles reporting on systematic quantitative evaluations of achievement, and in fact ESM shunned such articles. On one occasion the editor went so far as to dedicate an entire issue of the journal (vol. 6, no. 2, 1975) to a scathing critique of the North-American practice of using multiplechoice tests to measure mathematical achievement in general and of making international comparisons of achievement based on a common test in particular, as was done in the First International Mathematics Study (FIMS). Ironically, it was through this issue that ESM became especially well known in North America. Through its rst decade the journal continued, in the main, to publish essay articles that did not include any element of quantitative research. The table of contents for volume 4 (1973), for example, indicates that out of 33 articles, 16 discussed approaches to teaching various mathematics topics, nine were critical examinations of some psychological aspects of thinking about mathematics, two discussed linguistic variables that inuence mathematics learning, two were about computers and mathematics, two addressed student behaviour, one described Soviet research, and one traced the history of reforms in mathematics education in the UK.
79
In the late 1970s, however, the journal began to publish a wider variety of articles, including ones that reported on quantitative research relying on statistical analyses. The subject matter became more varied as well. While it was still true that most of the articles were essays on general ideas, the journal did publish reports of research on real classroom problems such as learning styles, grouping of students, gender differences, classroom interactions, and student attitudes and misconceptions. A great deal of this research, too, relied upon quantitative methods. As the journal became somewhat more closely associated with the International Group on the Psychology of Mathematical Education (PME), in that a great number of its editorial board members were also members of PME, it tended to publish manuscripts that reected the norms of research endorsed at the time by PME. Indeed, since the 1990s ESM has published two PME Special Issues, each of around 100 pages, every calendar year. These are explicitly identied as PME Special Issues, but are subject to the normal ESM reviewing process. Journal for Research in Mathematics Education (JRME) BEGINNINGS. The Journal for Research in Mathematics Education (JRME) is an American journal sponsored and managed by the National Council of Teachers of Mathematics (NCTM) that publishes, in English only, scholarly papers in mathematics education. Its rst issue appeared in January 1970 under the editorship of David C. Johnson of the University of Minnesota. It published ve research articles, two of which were Verbal and nonverbal assessment of the conservation of illusion-distorted length at the primary level, and Attitude changes in a mathematics laboratory utilizing a mathematics-through-science approach at the eighth-grade level. JRME editors are appointed by NCTM [Johnson, Romberg and Scandura 1994]. POLICY
STATEMENT.
The Editorial in the rst issue invited researchers to
[] submit papers which represent a coordinated approach to research on a major problem in learning, methodology, role of materials, etc. Such blocks of research could be published in a regular or special issue of JRME, or as an NCTM research monograph.
The journal was also willing to consider submissions of models and plans for large-scale research in an area, for publication in a special section called A forum for researchers. This policy statement has undergone a few changes over the years, reecting the changing wishes of the NCTM members as well
80
G. HANNA
as changes in the area of research in mathematics education. The most recent policy statement, posted on the Web, says that JRME is
devoted to the interests of teachers of mathematics and mathematics education at all levels (preschool through adult). [The JRME] is a forum for disciplined inquiry into the teaching and learning of mathematics. The editors encourage the submission of a variety of manuscripts : reports of research, including experiments, case studies, surveys, philosophical studies, and historical studies ; articles about research, including literature reviews and theoretical analyses ; brief reports of research ; critiques of articles and books ; and brief commentaries on issues pertaining to research.
SUBJECTS OF ARTICLES. In its early years JRME published mainly research reports. An examination of the subject index for volume 4 (1973), for example, indicates that of the 29 articles published in that year as many as 18 are research reports on the comparative effectiveness, as measured by achievement, of various class sizes and methods of teaching. Of the remaining eleven articles, those not concerned with measuring effects, four discuss psychological theories, four review various curricula and three report on other research on mathematics education or present responses to previous contributions. The titles of articles had a very practical sound in the rst decade of the journal. They almost always included the word effect (The effect of classsize on, The effect of organisers, The effects of instruction on) or the word comparison (A comparison of three strategies, A comparison of initially teaching), or the words correlating, determining and predicting. Indeed, throughout its rst ten years about two-thirds of the articles in JRME were experimental research reports that relied on statistical techniques such as analysis of variance or regression analysis. Even after the second decade such research reports still made up about ve-eighths of the articles [Lester and Lambdin 1998]. There had been some increase, however, in the proportion of articles that relied in whole or in part on qualitative methods, and that trend continued through the third decade. RESEARCH METHODS AND THEORETICAL FRAMEWORKS IN JRME AND ESM, 19902000. Over the last decade, a comparison of the research paradigms reported in JRME articles with those reported in ESM reveals more similarities than differences. Although JRME is still perceived, at least by Europeans, as a journal that stands out in not favouring articles on qualitative research, this perception is not borne out by the data. As shown in Figure 2, if one considers all articles, including those that do not report on research at all, the proportion
81
Qualitative Analysis Interviews Quantitative Analysis Empirical-analytic Observations Content Analysis Trend analysis Historical evidence Class discourse Longitudinal Study Descriptive Statistics Surveys Cross-cultural study CAI Hypothesis testing Action research Alternative Assessment Causal Modeling
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45
JRME ESM
FIGURE 2 Proportion of articles in JRME and ESM using each research paradigm
of articles that make use of quantitative analysis is indeed slightly higher in JRME than in ESM (28% vs. 18%), but the proportion of articles presenting qualitative analysis is the same in both journals. For the Learning of Mathematics (FLM) BEGINNINGS. This journal owes its existence, its unique character, and its prominence in the eld of mathematics education to its founder and rst editor, David Wheeler. It came into being in 1980 with the help of small grants from the Canadian Social Science and Humanities Research Council and from Concordia University in Montreal. David Wheeler was the editor of
82
G. HANNA
FLM for its rst 50 issues, which appeared between 1980 and 1997. In 1997, after Wheelers retirement, the Canadian Mathematics Education Study Group (CMESG) became its manager and appointed David Pimm as editor. FLM publishes papers in both English and French, though most appear in English. POLICY STATEMENT. From its inception, For the Learning of Mathematics (FLM) reected the broad interests and good judgment of David Wheeler. The name of the journal was intentionally chosen to emphasize the concern for learning and to acknowledge the inuence of Caleb Gattegno, known for his document For the Teaching of Mathematics [Gattegno 1963] and his other inuential contributions to mathematics education over four decades. FLM quickly established itself as a respected periodical that has consistently published articles reective of its policy statement :
[] to stimulate reection on and study of the practices and theories of mathematics education at all levels ; to generate productive discussion ; to encourage enquiry and research ; to promote criticism and evaluation of ideas and procedures current in the eld. It is intended for the mathematics educator who is aware that the learning and the teaching of mathematics are complex enterprises about which much remains to be revealed and understood.
In its suggestions to writers, FLM indicated that

Mathematics education should be interpreted to mean the whole eld of human ideas and activities that affect, or could affect, the learning of mathematics. [] The journal has space for articles which attempt to bring together ideas from several sources and show their relation to the theories or practices of mathematics education. It is a place where ideas may be tried out and presented for discussion.
A review of articles published in the early issues of FLM conrms that the editor had selected papers that stimulated reection on mathematics education, in line with these suggestions and with journal policy. Some characteristic titles were Communicating mathematics is also a human activity, Alternative research metaphors and the social context of mathematics teaching and learning, and When is a symbol symbolic ?. SUBJECTS OF ARTICLES. The articles published in FLM over the past two decades show a breadth of subject matter. They draw heavily on a number of disciplines not traditionally associated with mathematics education, such as sociology, linguistics, history, physics, arts, philosophy, and ethno-mathematics. Some indication of the place of such themes in FLM is given by the occurrence of certain key terms in the titles of the approximately 300 articles that appeared
83
in the years 19802000. Some variation of the word theory (theoretical, theories) appeared in 23 titles, history in as many as 20 titles, ethnomathematics in 12, language in 11, proof in 9, epistemology in 7, and philosophy in 7. FLM seems to have put a particular emphasis on articles that discussed the role of the history of mathematics in mathematics education (along with suggestions for its classroom use) and explored the concept of ethnomathematics and its potential application to mathematics education. In fact, FLM may be the rst scholarly journal of mathematics education to publish papers on these topics and in so doing to contribute signicantly to the development of these young areas of scholarly research. Another theme prominent in FLM is that of the nature of proof in mathematics and its implications for the teaching of proof. Another way to characterize FLM is to look at the kind of papers it did not publish. It has very few articles, only three in fact, that follow a traditional format of research papers in mathematics education : problem statement, review of relevant literature, rationale for the method of research and research results. There are no articles reporting on purely experimental classroom research and relying solely on statistical analyses. Some of the published articles do report on empirical research, but only as part of a wider discussion of some issue. No article appeared with a title such as The effect of X on Y. FLM has published relatively few purely empirical studies, and thus it is difcult to determine whether there was an editorial preference for any particular research method.
CONCLUSION The impressive number of active mathematics education journals is probably the most tangible indication that the discipline of mathematics education is alive and well. They have been very successful, collectively, in establishing and maintaining a wide variety of national and international forums for the publication and discussion of research on this topic. In so disseminating new information and encouraging debate, these journals also serve researchers as a valuable instrument of professional development.
84
G. HANNA
REFERENCES FURINGHETTI, F. [2003] Mathematical instruction in an international perspective : the contribution of the journal LEnseignement Math matique. This volume, 1946. e GATTEGNO, C. For the Teaching of Mathematics. Educational Solutions, New York, 1963. JACKSON, A. The slow revolution of the free electronic journal. Notices of the AMS (47) 9 (2000), 10531059. JOHNSON, D. C., T. A. ROMBERG and J. M. SCANDURA. The origins of JRME : a retrospective account. Journal for Research in Mathematics Education 25 (6) (1994), 560582. LESTER, F. and D. LAMBDIN. The ship of Theseus and other metaphors for thinking about what we value in mathematics education research. In : A. Sierpinska and J. Kilpatrick (eds.), Mathematics education as a research domain : a search for identity, 415425. Kluwer Academic Publishers, Dordrecht, 1998. SCHUBRING, G. (in cooperation with J. RICHTER). International Bibliography of Journals in Mathematical Education. (Schriftenreihe des IDM ; 23.) Univ. Bielefeld, Inst. f r Didaktik der Mathematik, Bielefeld, 1980. u UIPD. Ulrichs International Periodicals Directory (38th edition). Reed Elsevier, New Providence, NJ, 2000 (http://www.ulrichsweb.com/). ZDM. Zentralblatt f r Didaktik der Mathematik : International Reviews on Matheu matical Education. Fachinformationszentrum, Karlsruhe, 2000 (http://www.zkarlsruhe.de/z/publications/zdm/zdmzs.html ; also MATHDI [MATHematics DIdactics] : http://www.emis.de/MATH/DI.html).
REACTION by Jean-Pierre BOURGUIGNON
Five key points can be identied within the context of the changes that have occurred between 1900 and 2000, if we consider the key issues facing mathematics education today. THE INTERNATIONAL DIMENSION. Internationalism was at the very heart of the genesis of LEnseignement Math matique. Of course, mathematics then was e international, but there was a need to provide opportunities for communication. Today things have changed a great deal because of changes in the world we are living in rather than in mathematics itself. Travel is much easier, news is disseminated more swiftly, and altogether the way we feel about the world is very different. The internationalism then espoused by the journal is now commonplace. LINKS BETWEEN MATHEMATICS AND SOCIETY. Several articles in the early issues of LEnseignement Math matique address the responsibility of e mathematicians to educate, not only mathematicians, but also those engaged in other disciplines. That time coincided with the birth of many national societies. As far as France was concerned, a major stimulus for the creation of national scientic societies derived from the defeat of France by Germany in 1870 and the perceived view that France had not paid sufcient attention to the importance of science (in particular in its connections with engineering). Also in 1900 a big effort towards popularising science could be observed. Today the signicance of the link between mathematics and society (whose operation in a number of places is actually very dependent on sophisticated mathematics and mathematical methods) has not really percolated down to the teaching of mathematics (at least in France). It is essential to make much more explicit the link between the mathematics that is studied and the society
86
J.-P. BOURGUIGNON
it serves. In fact much of society is very critical of science in general and we see a decline in the number of students wishing to study science. It is very important to address this trend which comes in particular from a generally negative view of science. A second point to make about the link between society and mathematics is a change of the vocabulary used in mathematics today compared with the early issues of LEnseignement Math matique. An example is the use of the term e model which now appears in almost every book and curriculum statement about mathematics and was not there in the 1900s. This is more than fashion, and shows a conceptual change in the way we view mathematics. If we look at articles about geometry for example in the early issues of LEnseignement Math matique, we have the impression it is concerned with concrete objects. e The idea that mathematics represents things in an idealised way was not even present. Today this is very different and shows that this change of philosophy is now widely accepted. AN ARENA FOR CONFRONTATION. The early issues of LEnseignement Math matique were seen by its editors as an arena for the confrontation e of mathematicians and teachers of mathematics. The word confrontation is not pejorative but used here in the original sense of coming face to face with. The need still exists today but there are signicant differences, not least the development of a separate discipline of didactics, as Gert Schubring has noted. At that time the number of research mathematicians world-wide was about 200 in all ; today there are of the order of 60,000. The number of secondary school mathematics teachers today, world-wide, has become huge. Clearly a journal can no longer serve as a tool for the common interest of both groups (and, in any case, the journal is no longer playing this role). It is also noteworthy that there have been changes, at least recently, in the way teachers of mathematics and mathematicians perceive each other. To refer to the French situation, we can note that teachers of mathematics are no longer asking mathematicians to talk to them about ways of teaching mathematics (tricks to be used in the classroom) but are asking increasingly for information about the nature of new mathematics being done. This is a sign that a move towards both groups seeing each other as experts in their own areas is occurring. An important new development is the increasing use of computers, which is not to be seen in any way as a threat, as implied by the comments of Gert Schubring. But in order for it not to become one, it is vital that the rst experiences of students with computers in the secondary school be given by mathematicians and not by computer scientists. This means that
87
mathematics teachers need to become familiar with computers and computer software and make substantial efforts to incorporate this tool into their own teaching. This will have consequences on the relative weights of different areas of mathematics, the usual emphasis on analysis at the end of the school curriculum being challenged by theories and objects borrowed from discrete mathematics.
LENSEIGNEMENT MATHEMATIQUE AS A TOOL FOR INFORMATION FLOW. The journal saw itself as an important vehicle for the exchange of information. This is no longer a necessity. In fact, the need today is not creating a ow of information, but controlling it. This is a problem that still needs to be addressed at the right level, because we need to get the right information to the right people, and how this can be achieved is not so obvious. In any case, a paper journal is clearly no longer a suitable tool for the purpose of reaching teachers at all because of its too restricted circulation.
THE SCIENTIFIC LANDSCAPE. There has been a major change in the scientic landscape. Early issues of the journal did not really concern themselves with addressing other sciences. Today, as ever, mathematicians seem to be reluctant to talk enough and at the appropriate depth to other scientists. This attitude is extremely dangerous for mathematics, not only for opportunistic reasons, but also for the health of the discipline. One reason is that the scientic landscape itself has changed a great deal. Many new sciences have appeared (computer science has already been mentioned, but we could mention others) and the general balance between elds is now completely different, the dominating role and the now central importance achieved by the life sciences being one of the features to stress. The life sciences are attractive to politicians (because they address, and fortunately in several cases solve, important problems connected with the health of the population) and also attractive to students (because they offer new challenges and new frontiers to explore). In fact many new ideas for mathematics come from the sciences. To give one name, which brings us back to 1900, Henri Poincar was inspired by celestial mechanics and e theoretical physics. We need to look to sciences to provide stimulation for new mathematics. Science will demand new tools, and new mathematics will be generated although we dont know what it might turn out to be. We also need to involve other sciences in order to attract students to do mathematics. This has implications for the teaching of mathematics, which must expose students to other sciences as well as mathematics. Some countries train mathematics
88
J.-P. BOURGUIGNON
students by exposing them only to mathematics, which is not good for the students and also potentially weakens the development of the mathematical sciences. CONCLUDING REMARK. Initially LEnseignement Math matique provided e a tool by which mathematicians and mathematics teachers from different countries could compare their experiences, which were then very different. It is still the case today that experiences are very different in different parts of the world. Even if one considers just the European Union, one nds very diverse systems of mathematics education. If a unied platform is to be introduced within the EU, a task the governments have set themselves, then the education systems in the different countries need, at least, to be compatible. The tendency of bureaucracies everywhere when asked to provide a solution to such a problem is to go for a uniform solution. While compatibility is desirable, we stress that one should guard against this leading us into a wholly undesirable uniformity.
GENERAL DISCUSSION (reported by Chris WEEKS)
The discussion from the participants following J.-P. Bourguignons reaction to the three talks divided into two areas. On the one hand there were questions and comments about the journal LEnseignement Math matique itself e and on the other hand contributions to a discussion about the teaching of mathematics today, and particularly the importance of the key idea identied by J.-P. Bourguignon of Links with Society.
THE JOURNAL LENSEIGNEMENT MATHEMATIQUE. First, in answer to questions, Fulvia Furinghetti was able to provide some further information about the readership and dissemination of the journal. Although now almost entirely found in universities and other institutions, the journal in its early days was subscribed to by individual mathematicians and teachers of mathematics. Some of these appear to have been lone gures in countries like Rumania and Albania, but there was a relatively strong readership in France, Italy and Germany. Daniel Coray, on behalf of the present editorial board, reported that there are currently about 600 subscribers, although about 850 copies of each edition are printed to provide for a future time when new universities in developing nations, such as China, may wish to become subscribers and purchase previous copies. In addition to Europe and USA there is also a healthy readership in Japan. As to the content of the journal, it was noted that, in addition to technical mathematics articles, early editions contained articles relevant to the teaching of mathematics and articles about mathematics. The journal today consists almost entirely of research articles (some elementary, others not so elementary) and, notwithstanding its title, LEnseignement Math matique is not a journal e concerned with mathematics education. It is, however, a journal of ICMI and associated with IMU.
INTERNATIONAL DISTRIBUTION OF THE JOURNAL LENSEIGNEMENT MATHEMATIQUE (estimate based on shippings between 1996 and 2000)
Switzerland : 25 ; France : 45 ; Germany : 66 ; Austria : 7 ; Italy : 27 ; Belgium : 5 ; Netherlands : 40 ; Luxembourg : 1 ; Spain : 16 ; Portugal : 4 ; UK 1 ) : 42 ; Eire : 2 ; Denmark : 6 ; Norway : 3 ; Sweden : 7 ; Iceland : 1 ; Finland : 3 ; Croatia : 1 ; Slovenia : 1 ; Yugoslavia : 1 ; Czech Republic : 2 ; Poland : 7 ; Bulgaria : 1 ; Hungary : 5 ; Rumania : 3 ; Greece : 2 ; Russia : 4 ; Turkey : 1 ; Israel : 7 ; United Arab Emirates : 1 ; Tunisia : 2 ; South Africa : 2 ; Argentina : 3 ; Brazil : 8 ; Chile : 2 ; Mexico : 2 ; Uruguay : 1 ; Venezuela : 1 ; USA : 144 ; Canada : 19 ; Japan : 60 ; Korea : 1 ; China : 6 ; Hong Kong : 1 ; Taiwan : 1 ; Iran : 1 ; India : 9 ; Malaysia : 1 ; Singapore : 1 ; Vietnam : 2 ; Australia : 10 ; New Zealand : 2 Total 2 ) : 615
1 ) A number of copies sent to the UK or the Netherlands are in fact resent to other destinations (South America, Asia, Africa). The gures have therefore been corrected whenever some clear indications could be obtained from shippings in previous years, but they remain too high for these two countries. 2 ) This total includes about 100 copies which are exchanged with other periodicals from all parts of the world.
91
One important reason for the change in emphasis is the development of many national and international organisations now catering for the needs of teachers of mathematics, as well as researchers in the elds of psychology, history, didactics, etc. of mathematics, and these meet their client needs much better than a journal. A second reason is connected with the history of the journal, as explained further by Gert Schubring. From the beginning, LEnseignement Math matique should not be seen as e the only publication of ICMI. National reports on the state of mathematics education in different countries were solicited and these were also published. There were also ICMI Study Groups, whose reports were published. It should be noted, however, that contrary to the present ICMI the rst ICMI was not a permanent body : It was constituted by the 1908 International Congress of Mathematicians (ICM) with a mandate of four years, until the next Congress. Since this period proved to be too short, the mandate was extended in 1912 by four more years. Due to World War I, there was no Congress in 1916. Since there was, after the War, a ban against German scientists and since the Germans had been a highly active element within the ICMI work, the rst ICMI was dissolved in 1920. The rst again truly international Congress in the Inter-War period, 1928 in Bologna, reconstituted ICMI with a four-year mandate, extended at each following Congress until 1936 although the work never regained its former intensity. When ICMI was re-established in 1952, it became independent of the ICMs and attained a permanent character. The rst series of LEnseignement Math matique, until 1954, continued to provide e a forum for discussion about mathematics education but, for reasons given above, little of this appears in the second series. LINKS WITH SOCIETY. A number of participants were struck by la modernit de la probl matique of the founders of LEnseignement Math e e e matique. Certainly many of the problems facing the teaching of mathematics identied in the early issues of LEnseignement Math matique are present e today. A number of speakers commented on the importance of linking mathematics and mathematics teaching with society and this was also linked to the effectiveness and relevance of much of the teaching of mathematics today. While mathematics teaching of future mathematicians is by and large satisfactory, and is also generally satisfactory for users of mathematics such as physicists and engineers, mathematics education for the bulk of the population has serious weaknesses or worse quelle catastrophe as one speaker put it. Part of the cause lay in a certain detachment of mathematics education from
92
GENERAL DISCUSSION
the world around us, as if the processes of mathematics were in themselves sufcient to justify its existence. Many speakers picked up on the point made by J.-P. Bourguignon that mathematics education must be linked to other sciences. The sort of mathematics needed for the future non-professional user of the subject concerned a number of speakers. The day-to-day mathematical requirements for the citizen are decreasing while at the same time society is becoming more dependent on mathematics. In this sense, mathematics awareness is even more necessary but technical expertise is less necessary. Mathematics education needs to take account of this in terms of the content of the subject material and also its place in the curriculum. One danger is that mathematics could become like Latin, an optional subject for the interested scholar, and one speaker from Belgium suggested that this was already happening as a result of making it optional so that even chemistry and biology students no longer need to study mathematics. Other points made were that the learner is a person, with a life and experience outside the mathematics lesson. He or she is not a blank sheet on which new ideas and concepts can be inscribed, but comes to the mathematics lesson with many ideas (and perhaps many false ideas). Mathematics teaching that does not engage with the person is destined to fail. Linked to this is the point made by another speaker that, certainly for the majority of the population, mathematics must have a human face. In fact it would be a worthwhile venture for ICMI to set up a study group on this very topic.
GEOMETRY
` LA GEOMETRIE DANS LES PREMIERES ANNEES DE LA REVUE LENSEIGNEMENT MATHEMATIQUE Geometry in the rst years of the journal LEnseignement Math matique e by Rudolf BKOUCHE
The journal LEnseignement Math matique was closely associated with the reform e movement of scientic teaching which took place at the beginning of the 20 th century : the founding texts of the reform were in fact published in this journal. We use some of these texts to discuss a few questions raised by the renewal of geometry teaching. Understanding the global stakes of this reform implies taking into account the place that mathematics occupied in the physical and natural sciences and in technology at the turn of the 20th century. For the specic case of geometry, we must remember its dual character, at the crossroads of mathematics and physics, which conducted the reform movement through two major themes that have renewed the teaching of geometry. One is fusion that is, ignoring the traditional distinction between plane and solid geometry ; the other one is motion, more precisely the motion of solids in three-dimensional space. We will focus our attention on these two themes and on the way they appear in the texts published by LEnseignement Math matique. This will lead us to reconsider e the empiricist conception of geometry which marks these texts so strongly and which constituted an essential aspect of the whole period under review.
` LA GEOMETRIE DANS LES PREMIERES ANNEES DE LA REVUE LENSEIGNEMENT MATHEMATIQUE par Rudolf BKOUCHE
Se demander si un enfant a des dispositions ` pour la Math matique equivaut a se demander e sil en a pour l criture et pour la lecture. e [Laisant 1907]
INTRODUCTION On ne peut d tacher la revue LEnseignement Math matique de la r forme e e e de lenseignement scientique qui sest d velopp e au d but du XXe si` cle 1 ). e e e e e Cest dans la revue quont et publi s les textes fondateurs de la r forme et nous e e ` nous proposons dans cet expos daborder, a travers quelques textes publi s e e dans LEnseignement Math matique, les questions pos es par le renouvellement e e de lenseignement de la g om trie. e e Pour comprendre les enjeux g n raux de la r forme, il nous faut prendre e e e en compte la place des math matiques dans les sciences de la nature et dans e la technique au tournant des XIXeXXe si` cles et les conceptions positivistes e et empiristes qui marquent cette epoque. En ce qui concerne la g om trie il faut alors prendre en compte son e e caract` re mixte, au carrefour des math matiques et de la physique 2 ) ; cest e e ` ce caract` re mixte qui conduit la r forme a travers deux th` mes qui se e e e
1 ` ) Sur la r forme de lenseignement scientique en 1902, nous renvoyons a larticle de Bruno e ` Belhoste [1989] ainsi qu` notre article [Bkouche 1991a] et a louvrage Les Sciences au Lyc e a e [Belhoste, Gispert & Hulin 1996] 2 ) Geometry is a physical science ecrit W. K. Clifford [1885] dans un ouvrage, The Common Sense of the Exact Sciences, publi apr` s sa mort par Karl Pearson. A travers deux e e chapitres, Space et Position, Clifford explique comment la g om trie se rattache aux sciences e e de la nature.
` LA GEOMETRIE DANS LES PREMIERES ANNEES DE LA REVUE
97
proposent de renouveler lenseignement de la g om trie, dune part celui de la e e fusion, cest-` -dire labandon de la distinction traditionnelle entre la g om trie a e e plane et la g om trie dans lespace, dautre part celui du mouvement et plus e e particuli` rement du mouvement des corps solides. Cest donc sur ces deux e th` mes et la facon dont ils interviennent dans les textes de la revue que e ` nous centrerons notre expos . Cela nous conduira a revenir sur la conception e empiriste de la g om trie qui marque le tournant des XIXeXXe si` cles. e e e
LA
PLACE DES MATHEMATIQUES DANS LES SCIENCES DE LA NATURE ET DANS LA TECHNIQUE AU DEBUT DU XXe ` SIECLE
La r forme de lenseignement scientique, et particuli` rement de lenseie e gnement des math matiques, est marqu e par le succ` s de la math matisation e e e e des sciences de la nature, essentiellement des sciences physiques, et le r ole des math matiques dans le d veloppement des techniques. e e Cette math matisation pr sente deux aspects, dune part lexpression e e math matique des lois de la nature issue des travaux des physiciense math maticiens du XVIIe si` cle, dautre part, cons quence de cette math e e e e matisation, la possibilit accrue de pr diction des ph nom` nes, autant sur le e e e e plan qualitatif que sur le plan quantitatif. On pourrait citer trois grandes th ories e exemplaires, la m canique rationnelle, l lectromagn tisme et l nerg tisme. e e e e e Si nous ne pouvons entrer, dans le cadre de cet expos , dans une etude de e ces th ories, nous insisterons cependant sur le double aspect de leur succ` s : e e th orique dune part, technique dautre part. e Si la science math matis e prend cette importance dans le d veloppement e e e de la soci t , il devient n cessaire que la place des math matiques dans ee e e lenseignement soit r evalu e, autant dans lenseignement g n ral que dans e e e e lenseignement professionnel et la formation des ing nieurs. e De nombreux textes des premi` res ann es de LEnseignement Math matique e e e notent cette importance des math matiques dans la connaissance de la nature, e non seulement parce quelles mettent en forme la connaissance empirique mais aussi parce que le formalisme math matique est lui-m me cr ateur de e e e connaissance.
Les relations de la math matique avec le monde qui nous environne ne sont e donc pas accidentelles et articielles ; et si on oubliait ces relations, on ferait ` perdre a cette science son caract` re. e
98
R. BKOUCHE
ecrit Bettazzi [1900, 16] dans un article consacr a lapplication des e ` math matiques. On pourrait citer aussi des textes sur la m canique et son e e enseignement ([Laisant 1899], [Maggi 1901]), mettant en avant les liens entre la m canique th orique, laquelle rel` ve des math matiques, et la m canique e e e e e pratique et la n cessit dune th orisation. e e e Plus tard, lun des principaux animateurs de la r forme de 1902 en France, e Carlo Bourlet, expliquera, apr` s avoir insist sur la part de la connaissance de e e la nature dans le d veloppement des math matiques, combien il est juste que e e celles-ci, apr` s avoir pris leur autonomie, apportent aujourdhui leurs derniers e d veloppements aux sciences de la nature, remettant ainsi en question toute e distinction entre math matiques pures et math matiques appliqu es [Bourlet e e e 1910]. Cette insistance sur la part des math matiques dans la connaissance de la e nature posait une double question : dune part la n cessit dun enseignement e e des math matiques pour aborder l tude des sciences de la nature, dautre e e part la question de qui doit enseigner les math matiques : lenseignement des e math matiques doit-il etre pris en charge pas les enseignants des disciplines e qui en ont besoin ou doit-il etre assur par un corps sp cique denseignants e e de math matiques. e ` Cest a cette seconde question que sattaque Felix Klein dans un article de 1906. Sil reconnat la n cessit dun enseignement moins formel des e e math matiques dans les Hautes Ecoles Techniques (les ecoles ding nieurs), il e e pr cise combien une etude math matique pr alable peut faciliter la r solution e e e e ` de certains probl` mes pratiques. Cela le conduit a refuser que lenseignement e des math matiques soit laiss aux seuls praticiens [Klein 1906]. e e
LA
GEOMETRIE AU CARREFOUR DES MATHEMATIQUES ET DE LA PHYSIQUE
La position de la g om trie est particuli` re dans la mesure o` elle parle e e e u des objets de lespace, plus pr cis ment des corps solides ; en ce sens e e elle se rattache aux sciences de la nature. Il faut alors rappeler que la classique distinction entre g om trie rationnelle et g om trie pratique porte e e e e essentiellement sur la mani` re d tudier les propri t s g om triques des corps, e e ee e e mais la g om trie est une, du moins jusqu` la d couverte des g om tries e e a e e e non-euclidiennes. Cest la d couverte de la multiplicit des g om tries qui a e e e e ` conduit a distinguer une g om trie math matique, dont lexemple est donn e e e e par la construction hilbertienne [Hilbert 1899], et une g om trie physique qui e e
99
` sint resse a la structure de lespace dans lequel nous vivons, ce quEinstein e r sume de la facon suivante : e
` Pour autant que les propositions de la math matique se rapportent a la r alit , e e e elles ne sont pas certaines, et pour autant quelles sont certaines, elles ne se ` rapportent pas a la r alit 3 ). [Einstein 1921 (1972), 76] e e
Einstein pose alors la question des relations entre cette g om trie e e math matique et la g om trie physique, mais ce nest pas encore la quese e e tion de ce d but de si` cle qui sinterroge sur les liens entre les objets id aux e e e de la g om trie et les objets du monde r el. e e e Il semble cependant, en ce d but du XXe si` cle, que la d couverte des e e e g om tries non-euclidiennes ne soit pas encore comprise et accept e par tous, e e e comme le montre le d bat r current sur le postulat des parall` les dans les e e e premiers num ros de la revue. On peut alors noter que les tentatives de e d monstration du postulat des parall` les ou les critiques des arguments none e euclidiens sappuient sur des arguments dordre physique et limpossibilit de e penser un espace autre que leuclidien auquel nous sommes habitu s et qui e est devenu un cadre de pens e 4 ). e Ces r sistances encore fortes devant la possibilit dune g om trie none e e e euclidienne montrent que la distinction entre g om trie math matique et e e e g om trie physique nest pas faite malgr les ecrits des p` res de la g om trie e e e e e e non-euclidienne, Gauss, Lobatchevski et Bolyai, puis ceux de Riemann, de Klein ou de Poincar . Les mod` les euclidiens restent loin d tre convaincants ; e e e les g od siques de la pseudo-sph` re de Beltrami ne sont pas des droites e e e expliquent certains contradicteurs. Notons qu` la m me epoque (n du XIXe a e si` cle et d but du XXe) des arguments contre la g om trie non-euclidienne e e e e sont d velopp s par Cayley [1883] et Frege [1994]. e e En contrepoint, des auteurs vont publier des expos s de g om trie e e e e el mentaire en faisant ressortir la partie dicelle qui ne d pend pas du pose tulat des parall` les, en particulier la g om trie sph rique 5 ). Nous citerons e e e e ici louvrage de Dassen (math maticien argentin), Tratado elemental de Geoe
3 4
) Einstein distingue dans ce texte la g om trie axiomatique pure et la g om trie pratique. e e e e ) En un certain sens lespace euclidien est devenu une forme a priori de notre intuition, ce qui
pourrait etre une tentative dexplication des raisons qui ont conduit aux conceptions d velopp es e e par Kant dans son Esth tique transcendantale. e
5 ) Lind pendance de la g om trie sph rique par rapport au postulat des parall` les semble avoir e e e e e ` frapp les math maticiens dautant que cest a partir de consid rations de g om trie sph rique que e e e e e e Lobatchevski d veloppe les relations m triques et trigonom triques de sa g om trie [Lobatchevski e e e e e 1840].
100
R. BKOUCHE
metra Eucldea (tome I : g om trie plane, tome II : g om trie dans lespace), e e e e ouvrage longuement comment dans la revue 6 ). e Laspect physique de la g om trie joue un r le essentiel dans la mise en e e o place de la r forme et lon peut consid rer que cest via lenseignement de e e la g om trie que safrme dans lenseignement la conception empiriste des e e math matiques. e
LA
QUESTION DE LA FUSION
Notons dabord cette dissym trie du vocabulaire scolaire francais : on parle e de g om trie plane et de g om trie dans lespace, la premi` re fait r f rence e e e e e ee ` au plan, la seconde fait r f rence moins a lespace en tant que tel qu` ce qui ee a a lieu dans lespace ; si le plan est un objet g om trique que lon peut d nir, e e e lespace est le lieu de la g om trie, il ne fait que fournir les lieux que les e e ` corps occupent et remplissent , comme lexplique Euler dans ses Lettres a une Princesse dAllemagne [Euler 1768/1772]. Notons que langlais utilise les termes de plane geometry et de solid geometry 7 ). Il faut rappeler que, dans la g om trie grecque, si le plan est d ni comme e e e objet g om trique 8 ), la notion despace ny apparat pas ; on y etudie des corps, e e essentiellement des corps solides, repr sent s par des gures. e e La distinction entre la planim trie ( tude des gures planes) et la e e st r om trie ( tude des gures de lespace) proc` de de multiples raisons. ee e e e Dabord le r le sp cique que joue le plan dans les formes dexpression par o e lhomme de son rapport au monde, comme le montrent aussi bien le dessin que l criture, ensuite la st r om trie est une science difcile, comme lexplique e ee e d j` Platon au Livre VII [528 bd] de La R publique [1966, 287], enn, ea e dans le cadre de la rationalit grecque, le d veloppement logique exige que e e l tude des corps solides vienne apr` s celle des gures planes sur laquelle elle e e e sappuie, d veloppement qui est celui des El ments dEuclide o` la g om trie e u e e plane occupe la premi` re partie (Livres 14 et 6) alors que la st r om trie e ee e
6 ) Louvrage de Dassen fait lobjet de deux comptes rendus, le premier apr` s la publication e du tome I in LEnseign. Math. 7 (1905), 244246; le second apr` s la publication du tome II in e LEnseign. Math. 9 (1907), 7475. 7 ) Cependant en allemand on dit Geometrie der Ebene (g om trie du plan) et Geometrie des e e Raumes (g om trie de lespace); on dit aussi ebene Geometrie (g om trie plane) et r umliche e e e e a ` Geometrie (g om trie spatiale). Je dois ces renseignements a Nicolas Rouche. e e 8 ` ) Pour une revue des diverses d nitions de la notion de plan, nous renvoyons a louvrage e e de Heath [1908, Book I] sur les El ments dEuclide.
101
occupe les trois derniers livres (Livres 1113) ; ce d veloppement implique e un ordre des d monstrations, la d monstration dune propri t de g om trie e e ee e e plane ne saurait ainsi sappuyer sur des propri t s de st r om trie. Cest cet ee ee e ordre euclidien qui fonde la tradition g om trique occidentale (au sens large e e du terme, depuis le d veloppement scientique de l poque arabo-islamique e e jusqu` la r volution scientique de lEurope du XVIIe si` cle). Il faut noter a e e e cependant que les principes enonc s au d but du Livre I des El ments (pose e 9 tulats, axiomes) ne concernent pas la seule g om trie plane ). e e Nous ferons ici deux remarques. Dune part, la distinction euclidienne rel` ve dun ordre logique ; si les textes probl matiques, tels ceux dArchim` de, e e e prennent quelques libert s avec cet ordre, les textes dexposition se doivent de e le respecter. Dautre part, cette distinction ne se retrouve pas dans dautres aires culturelles, comme le montrent par exemple les math matiques chinoises qui e d veloppent directement ce quon pourrait appeler une m thode des volumes e e ` que lon peut comparer a la m thode des aires des math matiques grecques e e [Martzloff 1990]. Cette distinction sera remise en cause par le d veloppement de la g om trie e e e projective qui mettra en valeur ce que Chasles a appel lalliance intime et e ` syst matique entre les gures a trois dimensions et les gures planes [Chasles e ` 1837 (1989), 191]. Ce point de vue conduira a utiliser lespace pour etudier des propri t s planes. ee En 1826 Gergonne, etudiant la dualit des propri t s de situation et e ee remarquant comment celles-ci peuvent etre d duites ind pendamment du calcul e e ` ` et de la th orie des proportions a condition pour cela de passer tour a tour de e ` ` la g om trie plane a celle de lespace et de celle-ci a la premi` re [Gergonne e e e 1826], proposait d j` de revenir sur la classique division g om trie plane ea e e g om trie dans lespace, y compris dans lenseignement. e e A c t de ces raisons dordre th orique M ray, qui cite le texte de oe e e e Gergonne dans la pr face de l dition de 1874 de ses Nouveaux El ments e e de G om trie, avance une raison dordre pratique, pr cisant que la distinction e e e entre la g om trie plane et la g om trie dans lespace e e e e
est encore plus nuisible dans lenseignement professionnel, car la pratique des Arts r clame bien plus la connaissance approfondie des principales e combinaisons de droites et de plans, que celle de propositions th oriques e comme les propri t s des s cantes du cercle 10 ). [M ray 1874, xixii] ee e e
9 ) Notons que la d monstration du premier cas d galit des triangles (Livre I, proposition 4) e e e nimplique pas que les triangles soient dans un m me plan. De facon g n rale les cas d galit e e e e e des triangles ne sont pas des th or` mes de g om trie plane (cf. [Bkouche 2000]). e e e e 10
` ) Apr` s sa publication en 1874, louvrage de M ray conduira a quelques exp riences denseie e e
102
R. BKOUCHE
On peut rapprocher ce dernier argument, qui sera repris par les partisans de la r forme, de ceux d velopp s par Monge dans la pr face de sa G om trie e e e e e e descriptive [Monge 1798], v ritable plaidoyer en faveur de cette technique e nouvelle que repr sente, pour son inventeur, la g om trie descriptive. e e e ` Le d veloppement de la g om trie projective conduira a mettre en place e e e ce que lon a appel la fusion entre la planim trie et la st r om trie ; des e e ee e ouvrages seront publi s en France et en Allemagne au milieu du XIXe si` cle e e sans grand succ` s. e Louvrage de M ray sera red couvert par Laisant qui en exposera les e e principes dans LEnseignement Math matique en 1901 ; larticle de Laisant e [1901] sera suivi par dautres articles pr sentant les conceptions de M ray ; e e voir [Perrin 1903] ainsi que les articles de M ray lui-m me 11 ). e e Cest en Italie que lid e de la fusion se d veloppera avec le plus de succ` s e e e ` a la n du XIXe si` cle. En 1873, Luigi Cremona, soucieux de d velopper un e e e ` e enseignement el mentaire de g om trie projective publiera, a lusage des el` ves e e des Instituts Techniques, des Elementi di Geometria Projettiva [Cremona 1873]. Dans la pr face, Cremona insiste sur le caract` re technique de louvrage qui e e e ` doit conduire rapidement les el` ves a appliquer les connaissances th oriques e ` au dessin [1873 (1875), IX] ; quant a la fusion, Cremona explique :
D` s le commencement jalterne sans distinction les th or` mes de g om trie e e e e e plane avec ceux de la g om trie de lespace, parce que lexp rience ma e e e enseign , et dautres lont remarqu avant moi, que les consid rations de e e e lespace sugg` rent bien souvent le moyen de rendre facile et intuitif ce qui e ` e serait compliqu et difcile a d montrer par la g om trie plane. [Cremona 1873 e e e (1875), X]
Ainsi se d veloppera en Italie un courant fusionniste, deux ouvrages seront e ` publi s qui contribueront a populariser lid e de la fusion entre les deux e e g om tries, les Elementi di geometria de De Paolis en 1884, puis les Elementi e e ` di geometria de Lazzari et Bassani en 1891, correspondant a un enseignement effectivement donn a la Royale Acad mie de Livourne ; les id es fusionnistes, e` e e telles quelles se sont d velopp es en Italie, font lobjet de deux articles de e e la revue, lun de Candido [1899], lautre de Loria [1905].
` gnement dans lAcad mie de Dijon (M ray etait professeur a la Facult des Sciences de Dijon) e e e non sans quelque r ussite, mais ces exp riences sarr teront devant lindiff rence, sinon lhostilit , e e e e e de lAdministration (!) et la m ance de certains enseignants, comme lexplique M ray lui-m me e e e dans la pr face de la seconde edition [M ray 1903] de son ouvrage (Jobard, Dijon). e e
11 ) [M ray 1901]; cet article est une partie dun article plus ancien publi dans la Revue e e bourguignonne de lenseignement sup rieur, 1892. Lors de la seconde edition de ses Nouveaux e e el ments de g om trie, Charles M ray publiera un article de pr sentation de louvrage [M ray e e e e e 1904].
103
Ce rapprochement, dans lenseignement, de la g om trie plane et de la e e g om trie dans lespace ne sera pas sans poser probl` me ; ainsi Hadamard, e e e e dans la pr face de la seconde edition de ses Lecons de g om trie el mentaire, e e e publi e a l poque de la r forme, refusera de fondre la g om trie plane et e ` e e e e la g om trie dans lespace , et il expliquera : e e
Que cette fusion soit pr f rable au point de vue logique, je le veux bien. Mais ee ` il me parat que, p dagogiquement, nous devons penser tout dabord a diviser e les difcult s. Celle de voir dans lespace en est une s rieuse par elle-m me, e e e que je ne consid` re pas comme devant etre ajout e tout dabord aux autres. e e [Hadamard 1906]
LA
QUESTION DU MOUVEMENT
Ici encore les r formateurs du d but du XXe si` cle vont remettre en cause e e e une tradition bien etablie depuis la g om trie grecque. e e ` Les raisons de l limination du mouvement sont a la fois dordre e m taphysique et dordre scientique ; mais peut-on distinguer science et e m taphysique lorsque lon parle de la pens e grecque ? Les paradoxes de e e Z non ont montr les difcult s de parler du mouvement et lon peut penser e e e que devant ces difcult s, mieux valait circonscrire le domaine de la connaise ` sance rationnelle a ce qui peut etre objet de discours. Cest lune des raisons qui ` conduit Platon a proclamer : , , ` , 12 ) [La R publique, Livre VII, 526e]. ` e On peut alors consid rer que la force de la construction euclidienne vient e de ce quelle elimine tout recours au mouvement dans son discours alors que la physique aristot licienne a c d la place lorsque le temps est devenu une e e e grandeur g om trique et dune certaine facon statique. Dans cette nouvelle e e e physique g om tris e le mouvement echappe aux paradoxes des El ates pour e e e devenir objet de science, cest-` -dire objet dun discours rationnel. a Pourtant le mouvement est pr sent dans la g om trie grecque, dabord avec e e e le principe de l galit par superposition (laxiome 4 ou 8 suivant les editions e e e des El ments dEuclide), principe que nous enoncons dans la traduction de Ho el [1867, 13] : les grandeurs que lon peut faire concider lune avec u lautre sont egales entre elles , principe fondateur de la g om trie dans la e e mesure o` cest ce principe qui permet de comparer entre eux les objets u
12 ` ) Si [la g om trie] oblige a contempler lessence, elle nous convient; si elle sarr te au e e e devenir, elle ne nous convient pas. [Platon 1966, 285]
104
R. BKOUCHE
g om triques et donc de les mesurer. Mais cette intervention du mouvement e e est vite elimin e, et cest le r le du principe de l galit par superposition e o e e que de permettre cette elimination ; en effet, ce principe une fois enonc , e le probl` me est de trouver des crit` res d galit a priori qui permettent e e e e de se dispenser de la superposition effective, ce seront les classiques cas d galit des triangles, lesquels sont, en ce sens, fondateurs de la rationalit e e e g om trique euclidienne [Bkouche 1991b ; 2000]. Dans le cadre de cette e e rationalit , et conform ment au dogme platonicien rappel ci-dessus, lappel e e e ` a des consid rations de mouvement est interdit en droit dans le raisonnement e g om rique. e e ` Cest la g om trisation du temps qui va conduire a la possibilit dune e e e ` etude scientique du mouvement et en retour a son utilisation pour r soudre e certains probl` mes de g om trie, ainsi le probl` me des tangentes 13 ). Cette e e e e ` e g om trisation du mouvement conduira a d nir au XIXe si` cle une g om trie e e e e e du mouvement ou g om trie cin matique, g om trie quAm d e Mannheim, e e e e e e e ` dans son cours a lEcole polytechnique, d nit ainsi : e
La Cin matique a pour objet l tude du mouvement ind pendamment des forces ; e e e la G om trie cin matique a pour objet l tude du mouvement ind pendamment e e e e e des forces et du temps, cest-` -dire quelle a pour objet l tude des d placements. a e e Nous r servons lexpression de d placement pour un mouvement dans lequel e e on ne consid` re pas la vitesse. [Mannheim 1880, treizi` me lecon] e e
Ce nest pas ici le lieu de d velopper l tude de cette histoire mais nous e e ` citerons Jules Ho el qui, ouvrant ainsi la voie a un renouvellement de lenseiu gnement de la g om trie, ecrivait en 1867, dans son Essai critique sur les e e e principes fondamentaux de la g om trie el mentaire : e e
Cest par suite dune confusion did es que plusieurs g om` tres veulent e e e e bannir des el ments de g om trie la consid ration du mouvement. e e e Lid e du mouvement, abstraction faite du temps employ a laccomplir, e e ` cest-` -dire lid e du mouvement g om trique, nest pas une id e plus complexe a e e e e que celle de grandeur ou d tendue. On peut m me dire, en toute rigueur, que e e cette id e est identique avec celle de grandeur, puisque cest pr cis ment par e e e ` le mouvement que nous parvenons a lid e de grandeur. e Ce mouvement g om trique, quune equivoque de langage a fait confondre e e avec le mouvement dans le temps, objet de la cin matique, ne peut pas d pendre e e dune autre science que la g om trie pure. e e Il est avantageux dintroduire cette id e de mouvement g om trique le plus e e e t t et le plus explicitement possible. On y gagne beaucoup sous le rapport o de la clart et de la pr cision du langage, et lon se trouve mieux pr par a e e e e ` introduire plus tard dans le mouvement les notions de temps et de vitesse.
13 ) On pourrait citer les travaux de Roberval en 1636 sur la d termination des tangentes, e travaux publi s en 1693 [Roberval 1693]. e
105
` Cest dailleurs ce que tous les auteurs font a leur insu ; et il serait difcile de trouver une seule d monstration dune proposition fondamentale de la g om trie, e e e dans laquelle nentre pas lid e de mouvement g om trique, plus ou moins e e e d guis e. [Ho el 1867, Note II] e e u
` Cest ce courant did es qui permettra a Charles M ray dutiliser le e e e mouvement dans la construction de la g om trie el mentaire via les translations e e et les rotations ; aux translations sont alors associ es les notions de droite et e de parall lisme, aux rotations sont associ es la notion de perpendiculaire ainsi e e que les notions dangle et de cercle. Cette transgression de la tradition euclidienne sera diversement appr ci e e e ` et on reprochera aux partisans de la r forme un manquement a la rigueur dans e le raisonnement. Il est vrai que la construction de M ray est assez compliqu e e e ` et, par cela-m me, loin d tre a labri des critiques ; sa recherche dun enseie e gnement plus intuitif et le souci conjoint de pr server une certaine rigueur e de lexpos explique cette complication. Mais derri` re la complication du e e texte de M ray on voit apparatre les id es dont je viens de parler et surtout, e e mais cest peut- tre le point essentiel de la pol mique, laspect exp rimental e e e de la g om trie. Dautres ouvrages paratront au moment de la r forme qui e e e d velopperont ce point de vue tels ceux de Carlo Bourlet [1906/1908 ; 1912] e ou celui dEmile Borel [1905]. Il faudrait citer aussi la g om trie de lajustage d velopp e par Jules e e e e ` Andrade, professeur a lUniversit de Besancon, o` il a propos en 1905 un e u e cours de chronom trie. Comme il lexplique [Andrade 1905], il destinait un e tel enseignement aux ls des patrons dhorlogeries ; en fait il rencontre un public douvriers, ce qui lam` ne a penser un cours de g om trie permettant e ` e e dacc der aux principales connaissances sans avoir suivi le cursus classique. e ` Cest loccasion pour lui de penser lenseignement de la g om trie a travers e e les pratiques de mouvements utilis es en atelier. Il construit ainsi son cours en e termes de translations et de rotations, moins comme d nitions formelles mais e comme descriptions de mouvements. On peut alors introduire les parall` les via e le mouvement de translation ; par ailleurs l tude du mouvement dun solide e ayant deux points xes permet une d nition empirique de la droite comme e lensemble des points xes dans le mouvement dun solide ayant deux points ` xes 14 ). Cela le conduit a construire une g om trie naturelle [Andrade 1908], e e dont il explique quelle est fond e cin matiquement. e e
14 ) Notons que cette d nition est donn e par Leibniz dans sa Caract ristique g om trique e e e e e [Leibniz 1995, 161]. Cette remarque pose la question de la relation complexe entre le rationalisme leibnizien et la connaissance empirique.
106
R. BKOUCHE
GEOMETRIE
ET DESSIN
Les r formateurs ont mis laccent sur lenseignement du dessin g om trique ; e e e un tel enseignement a un double role, dune part son importance dans lapprentissage de la g om trie, les gures apparaissant du moins dans e e la premi` re etude de la g om trie comme les objets de la g om trie, e e e e e ` dautre part sa place dans lenseignement professionnel ou la g om trie dee e scriptive joue un r le important comme mode de repr sentation des objets o e de lespace. On peut noter ici le nombre important de trait s de g om trie e e e descriptive publi s dans plusieurs pays europ ens et recens s par la revue. e e e Cest en partie pour assurer la liaison entre le dessin g om trique et e e lenseignement de la g om trie quEmile Borel proposera en 1904 la mise e e en place de laboratoires de math matiques, point sur lequel nous reviendrons e ci-dessous.
UNE
CONCEPTION EMPIRISTE DE LA GEOMETRIE
La r forme du d but du XXe si` cle est marqu e par lempirisme ; les e e e e math matiques sont le moyen de d crire la nature, mais si elles ont cette e e propri t cest quelles sont issues de connaissances naturelles comme le ee ` rappelle Bourlet [1910]. Quant a la g om trie, en tant quelle est science des e e corps solides [Poincar 1902 (1968), 86], elle participe autant des sciences de e la nature que des sciences math matiques. On met ainsi laccent sur la part e de physique qui sous-tend la g om trie. e e Nous citerons ici un article publi dans le premier volume de la revue dans e la mesure o` il apparat comme un manifeste de cette conception empiriste de u la connaissance, savoir, larticle de Laurent [1899], article que lon peut situer ` a la fois dans la tradition des Lumi` res (de dAlembert [1759] et Condillac e ` [1746] a Lacroix [1804]) et dans la tradition du positivisme comtien. Lorigine de nos connaissances se trouve dans nos sens , ecrit ainsi lauteur [Laurent 1899, 383] expliquant les trois modes de connaissance : lobservation, lexp rience (nous dirions aujourdhui lexp rimentation), le e e raisonnement, et il pr cise : toute science passe par trois phases suce cessives : 1o la phase dobservation, 2o la phase de raisonnement, 3o la phase exp rimentale [Laurent 1899, 385]. e On peut remarquer lordre des phases : la phase exp rimentale y apparat e moins comme une phase d laboration de la connaissance que comme une e
107
phase de v rication 15 ), la seconde place, apr` s la phase empirique d nie e e e par lobservation, etant celle du raisonnement. La place accord e au raisonnement montre la place de labstraction dans e lactivit scientique, abstraction que lauteur d nit comme une convention e e que lon fait de n gliger toutes les propri t s dun objet, pour ne tenir compte e ee que dune seule dentre elles [Laurent 1899, 393]. On peut comparer cette e conception avec celle de dAlembert qui ecrit dans son Essai sur les el ments de philosophie :
Labstraction en effet nest autre chose que lop ration par laquelle nous e consid rons dans un objet une propri t particuli` re, sans faire attention aux e ee e autres. [dAlembert 1759 (1986), 29]
Laurent propose alors une classication des sciences qui sappuie sur la plus ou moins grande part de la connaissance sensible qui intervient dans l laboration de cette science ; il place en t te la Math matique, quil r duit e e e e ` a la science des nombres (arithm tique et alg` bre) en ce que le r le des sens e e o y est presque nul, m me sil consid` re que la notion de nombre est issue des e e activit s li es au comptage. e e La G om trie est alors consid r e par Laurent comme une science physique e e ee parce quelle emprunte au t moignage des sens la notion despace et la notion e de d placement [Laurent 1899, 398], Laurent distinguant alors les notions de e ` d placement et de mouvement, cette derni` re notion faisant appel a la notion de e e temps. Mais si la g om trie commence comme une science dobservation, cest e e le raisonnement qui permet son d veloppement en apportant les moyens de e d couvrir de nouvelles propri t s des gures. De fait lentr e de la g om trie e ee e e e dans la math matique pure est li e a la g om trie analytique consid r e comme e e ` e e ee une reconstruction num rique de la g om trie ; cest une telle reconstruction e e e qui permet de la situer dans une science plus g n rale, la Pang om trie 16 ), e e e e g n ralis e elle-m me avec linvention des g om tries multidimensionnelles. e e e e e e Cest une fois cette construction rendue possible que la g om trie devient e e exp rimentale. e Lordre des phases placant le raisonnement avant lactivit est un point im e portant de larticle de Laurent qui dune part permet d viter un r alisme naf et e e dautre part permet de distinguer lobservation empirique et lexp rimentation, e cette derni` re supposant une activit pr alable de raisonnement. e e e
15 ) Cela remet en place certains jugements h tifs sur lempirisme et montre le r le du a o raisonnement dans la conception empiriste de la connaissance. 16 ` ) Rappelons que pang om trie est le nom que Lobatchevski donne a sa g om trie g n rale e e e e e e ` qui contient a la fois la g om trie euclidienne et la g om trie non-euclidienne; cf. [Lobatchevski e e e e e 1855] : ce texte a et publi simultan ment en russe et en francais en 1855. e e
108
R. BKOUCHE
` Le raisonnement, quant a lui, sappuie sur lobservation comme lexpliquera Laisant dans un article ult rieur : Il est certain que les raisonnements ne e peuvent tenir lieu de la connaissance des faits. [Laisant 1902, 8] Cest encore lempirisme qui guide M ray lorsquil explique dans son e article sur lenseignement des math matiques : Entre lexactitude dune e hypoth` se physique et l vidence dun axiome math matique, il ny a donc e e e aucune diff rence essentielle. [M ray 1901, 178179] e e Il pr cise dans ce m me article comment la g om trie sinscrit, moins dans e e e e les math matiques pures que dans les math matiques appliqu es. Apr` s avoir e e e e pr sent les deux grands chapitres des math matiques, lAnalyse math matique, e e e e science g n rale des nombres, et la G om trie, qui etudie la forme et l tendue e e e e e des corps ainsi que leurs positions relatives, il ecrit :
Lusage a r uni ces deux sciences sous la d nomination g n rique de Math e e e e e matiques pures, et a group dans les Math matiques appliqu es, toutes celles e e e r sultant effectivement de lapplication de lAnalyse et de la G om trie au e e e d veloppement dun tr` s petit nombre de notions sp ciales. Il semblerait plus e e e ` naturel de mettre a part lAnalyse, qui pr te sans cesse ses principes mais e nen emprunte aucun ailleurs, de r server le nom de Math matiques pures e e ` ` a ses diverses branches, puis de placer a sa suite, dans les Math matiques e appliqu es, toutes les sciences trouvant dans ses formules un appui essentiel et e ` continu. En les rangeant dans lordre ou chacune est n cessaire aux suivantes e mais non aux pr c dentes, on y rencontrerait la G om trie, la M canique, la e e e e e Physique math matique, . [M ray 1901, 173] e e
e Plus tard, dans la pr face de la seconde edition de ses Nouveaux el ments e de g om trie il parlera de faits g om triques et de la vision des faits de e e e e lespace [M ray 1903, vivii]. e e Mais la g om trie est marqu e par le fait davoir et la premi` re science e e e e rationalis e, cest-` -dire la premi` re a r pondre aux exigences des Seconds e a e ` e Analytiques dAristote. De ce fait la part de connaissance empirique de e la g om trie a et masqu e par le discours rationnel qui la soutenait. La e e e e m canique ou l lectromagn tisme ont et rationalis s beaucoup plus tard et e e e e leur caract` re empirique reste pr sent. e e Borel est conscient de cela lorsquil propose des travaux pratiques de g om trie dans lenseignement, lors de la conf rence de 1904 dont nous avons e e e d j` parl [Borel 1904], conf rence qui r sume ses conceptions p dagogiques. ea e e e e Laspect exp rimental de la g om trie apparat aussi avec les descriptions et e e e lusage dinstruments g om triques. Nous pourrions citer en France louvrage e e de Fourrey [1907], qui consacre un chapitre aux instruments g om triques. e e Nous pourrions aussi signaler une s rie darticles publi s dans LEnseignement e e Math matique sur la st r oscopie. e ee
109
e ` Cela nous renvoie a une lecture empiriste des El ments dEuclide que lon ne peut d velopper ici. En ce sens les propositions de r forme de lenseignee e ment de la g om trie restent proches du point de vue euclidien m me si elles e e e en renouvellent le style en mettant laccent sur lorigine empirique des objets de la g om trie. e e
QUELQUES
REMARQUES EN CONCLUSION
Nous voudrions terminer cet article en parlant de lactualit de certains e articles autour de la r forme du d but du XXe si` cle. e e e Outre le fait que certains probl` mes sont toujours actuels, en particulier e celui de d cider qui enseigne les math matiques dans lenseignement profese e sionnel, la question des rapports des math matiques avec les autres disciplines e reste permanente. Il faudrait alors revenir sur laspect exp rimental quil est n cessaire de e e replacer, comme le fait larticle cit de Laurent, dans un cadre gobal. En e particulier la part dempirisme de la g om trie ne saurait etre d tach e de e e e e la part rationnelle. Cest en ce sens quil faut comprendre la proposition de laboratoires de math matiques propos e par Emile Borel. Mais cela demande e e aussi de retrouver la part dempirisme de la g om trie grecque. e e Cest la prise en compte de cette part dempirisme qui fait d faut dans e lenseignement daujourdhui et, en ce qui concerne la g om trie, loubli e e quelle est aussi une science physique. Peut- tre parce que m me si lon e e prend en compte lapport des math matiques aux sciences de la nature, pour e ne parler que delles on a oubli que cet apport est possible uniquement e parce que les math matiques (m me si elles sen sont d tourn es pour mener e e e e leur vie propre) se sont construites sur des connaissances empiriques. Le travail cit de Jules Andrade construisant ce quil appelle une e g om trie naturelle nest pas seulement un artefact p dagogique utilis e e e e ` parce quil sadressait a des ouvriers ne poss dant pas la culture g om trique e e e classique ; en introduisant explicitement le mouvement dans son cours de g om trie il effectuait un retour aux origines et cest via ce retour aux orie e ` gines quil pouvait esp rer conduire a une meilleure compr hension de la e e e g om trie el mentaire. e e En ce sens, la lecture des textes des premiers num ros de la revue peut e etre un salutaire rappel pour lenseignement daujourdhui.
110
R. BKOUCHE
BIBLIOGRAPHIE e J. [1759] Essai sur les el ments de philosophie, ou sur les principes des connaissances humaines. Fayard, Paris, 1986. ANDRADE, J. Lenseignement scientique aux ecoles professionnelles et les math mae tiques de ling nieur. LEnseign. Math. 7 (1905), 2127. e Le premier livre de la g om trie naturelle. LEnseign. Math. 10 (1908), 185207, e e 296318, 391411. BELHOSTE, B. Les caract` res g n raux de lenseignement secondaire scientique de la e e e ` n de lAncien R gime a la Premi` re Guerre mondiale. Histoire de lEducation e e 41 (janvier 1989), 346. BELHOSTE, B., H. GISPERT et N. HULIN (eds.) Les sciences au lyc e. Vuibert, Paris, e 1996. BETTAZZI, R. Lapplication dans lenseignement de la math matique. LEnseign. Math. e 2 (1900), 1430. BKOUCHE, R. [1991a] Variations autour de la r forme de 1902/1905. In : H. Gispert, e La France math matique, 181213. Cahiers dHistoire et de Philosophie des e Sciences No 34, Paris, 1991. [1991b] De la g om trie et des transformations. Rep` resIREM 4 (juil. 1991), e e e 134152. Quelques remarques autour des cas d galit des triangles. Bulletin de lAPMEP e e 430 (sept.oct. 2000). BOREL, E. Les exercices pratiques de math matiques dans lenseignement secondaire. e Conf rence faite le 3 mars 1904 au Mus e p dagogique. In : uvres, t. 4, 2225 e e e 2256. Editions du C.N.R.S., Paris, 1972. G om trie. Armand Colin, Paris, 1905. e e BOURLET, C. Cours abr g de g om trie (2 tomes). Hachette, Paris, 1906/1908. e e e e La p n tration r ciproque des math matiques pures et des math matiques ape e e e e ` pliqu es dans lenseignement secondaire (conf rence a la r union de la Come e e mission internationale de lenseignement math matique). LEnseign. Math. 12 e (1910), 372387. e El ments de g om trie. Hachette, Paris, 1912. e e CANDIDO, G. Sur la fusion de la planim trie et de la st r om trie dans lenseignement e ee e e de la g om trie el mentaire en Italie. LEnseign. Math. 1 (1899), 204215. e e CAYLEY, A. Presidential address to the British Association. Report of the British Association for Advancement of Science (September 1883), 337 ; no. 784 in Collected Mathematical Papers, Vol. XI, 429459. CHASLES, M. [1837] Apercu historique sur lorigine et le d veloppement des m thodes e e en g om trie. Gabay, Paris, 1989. e e CLIFFORD, W. K. The Common Sense of the Exact Sciences, ed. by Karl Pearson. Appleton, New York, 1885. (Nombreuses r editions ; en particulier : Knopf, New e York, 1946 ; Dover, New York, 1955.) CONDILLAC, Abb de [1746] Essai sur lorigine des connaissances humaines. Editions e Galil e, Paris, 1973. e e CREMONA, L. [1873] Elementi di geometria projettiva. Torino, 1873 ; El ments de g om trie projective (traduction Dewulf). Gauthier-Villars, Paris 1875. e e
DALEMBERT,
111
EINSTEIN, A. [1921] La g om trie et lexp rience. In : R exions sur l lectroe e e e e dynamique, l ther, la g om trie et la relativit , textes traduits par M. Solovine e e e e et M.-A. Tonnelat, Gauthier-Villars, Paris, 1972. ` EULER, L. [1768/1772] Lettres a une princesse dAllemagne. Charpentier, Paris, 1859. FOURREY, E. Curiosit s g om triques. Vuibert, Paris, 1907. e e e FREGE, G. Sur la g om trie non-euclidienne (18991906). In : Ecrits posthumes, traduits e e de lallemand sous la direction de Ph. de Rouilhan et Cl. Tiercelin, 199201. Editions J. Chambon, Nmes, 1994. e GERGONNE, J. D. Consid rations philosophiques sur les el ments de la science de e l tendue. Annales de Math matiques XVI (1826), 209231. e e e HADAMARD, J. Lecons de g om trie el mentaire (2 volumes). Colin, Paris, 1898/1901. e e (2e edition : 1906.) HEATH, Th. L. [1908] Euclid : The Thirteen Books of the Elements (3 volumes), translation and commentaries. Dover, New York, 1956. HILBERT, D. [1899] Les fondements de la g om trie, edition critique pr par e par Paul e e e e Rossier. Dunod, Paris, 1971. e HOUEL, J. Essai critique sur les principes fondamentaux de la g om trie el mentaire. e e Gauthier-Villars, Paris, 1867. KLEIN, F. De lenseignement des Sciences math matiques et physiques dans les e Universit s et les Hautes Ecoles techniques. LEnseign. Math. 8 (1906), 525. e ` e LACROIX, S.-F. R exions sur lordre a suivre dans les el mens de g om trie, sur e e e e la mani` re de les ecrire, et sur la m thode en math matiques. In : El mens de e e e G om trie, quatri` me edition. Courcier, Paris, 1804. e e e LAISANT, C. A. La m canique rationnelle et la m canique appliqu e. LEnseign. Math. e e e 1 (1899), 237246. Une exhumation g om trique. LEnseign. Math. 3 (1901), 98105. e e A propos dun discours de M. Blutel [` la distribution des prix du Concours a g n ral]. LEnseign. Math. 4 (1902), 69. e e La math matique, philosophie, enseignement. 2e ed., Gauthier-Villars, Paris, 1907. e LAURENT, H. Les principes fondamentaux des connaissances humaines. LEnseign. Math. 1 (1899) 381419. LEIBNIZ, W. G. Caract ristique g om trique, texte etabli, introduit et annot par e e e e J. Echeverra ; traduit, annot et postfac par M. Parmentier. Mathesis , Vrin, e e Paris, 1995. LOBATCHEVSKI, N. I. [1840] Etudes g om triques sur la th orie des parall` les, traduit e e e e de lallemand par J. Ho el, M moires de la Soci t des Sciences Physiques et u e ee Naturelles de Bordeaux 4 (1866), 83120. [1855] Pangeometry. English translation from French. In : D. E. Smith, A Source Book in Mathematics, 360374. McGraw-Hill, New York, 1929 (et aussi Dover, New York, 1959). e LORIA, G. Sur lenseignement des math matiques el mentaires en Italie. LEnseign. e Math. 7 (1905), 1120. MAGGI, G.-A. R exions sur lexposition des principes de la m canique rationnelle. e e LEnseign. Math. 3 (1901), 240261. MANNHEIM, A. Cours de g om trie descriptive de lEcole polytechnique. Gauthiere e Villars, Paris, 1880.
112
R. BKOUCHE
MARTZLOFF, J.-C. Quelques exemples de d monstrations en math matiques chie e noises. In : La d monstration math matique dans lhistoire (colloque Inter-IREM e e Besancon). IREM, Lyon, 1990. e MERAY, Ch. Nouveaux el ments de g om trie. Savy, Paris, 1874. Nouvelle edition revue e e et augment e : Jobard, Dijon, 1903. e Lenseignement des math matiques. LEnseign. Math. 3 (1901), 172194. e e Justication des proc d s et de lordonnance des Nouveaux el ments de g om trie e e e e par lauteur. LEnseign. Math. 6 (1904), 89123. MONGE, G. [1798] G om trie descriptive. 4e ed., Courcier, Paris, 1820. e e PERRIN, E. La m thode de M. M ray pour lenseignement de la g om trie. LEnseign. e e e e Math. 5 (1903), 441446. PLATON. La R publique, traduction Baccou. Garnier Flammarion, Paris, 1966. e POINCARE, H. [1902] La science et lhypoth` se, pr face de J. Vuillemin. Flammarion, e e Paris, 1968. ROBERVAL, G. P. Observations sur la composition des mouvemens, et sur le moyen de trouver les touchantes des lignes courbes. In : Divers ouvrages de math matique e et de physique, par Messieurs de lAcad mie Royale des Sciences. Paris, 1693. e Et aussi in : M moires de lAcad mie Royale des Sciences, tome VI. Paris, 1730. e e
GEOMETRY : 195070 ` La g om trie de 1950 a 1970 e e par Geoffrey HOWSON
` Les deux d cennies de 1950 a 1970 ont une importance particuli` re, non seulement e e parce que lenseignement de la g om trie dans les ecoles y faisait lobjet de e e passablement de discussions (ce qui na rien dexceptionnel, puisque ce d bat est e toujours dactualit ), mais surtout en raison du foisonnement dactivit s qui eurent lieu e e effectivement dans les classes. De plus, les exp rimentations vari` rent enorm ment e e e dans leurs objectifs, les conditions de leur mise en uvre et leurs r sultats. Et il ny e ` avait pas de consensus sur les directions a suivre. Or, peu de traces subsistent de cette p riode de fertile instabilit , et celles qui restent ne permettent pas de bien e e comprendre ce que les auteurs de ces innovations se proposaient dobtenir. Nous aborderons bri` vement les questions suivantes : Quest-ce qui provoqua la e ` r volution dans lenseignement de la g om trie ? Que cherchait-on a faire ? Qua-t-on e e e obtenu ? Qua-t-on appris ? (Ou plut t : Quelles lecons aurait-on pu tirer, m me si cela o e e na souvent pas et fait ?) Nous remarquons quau d but de cette p riode il ny avait aucun point de vue e e universellement partag sur les buts de lenseignement de la g om trie. Plusieurs e e e traditions diff rentes et linuence exerc e par les math maticiens des universit s (qui e e e e avaient aussi une opinion sur les math matiques quil fallait enseigner dans les ecoles) e ont toutes jou des r les consid rables pour d nir la nature des r formes accomplies e o e e e dans les diff rents pays. e En particulier, on remarque des clivages entre des pr sentations strictement e axiomatiques et dautres plus pragmatiques ; ou entre celles qui se raccrochaient encore ` a une approche euclidienne (mais avec de nouveaux syst` mes daxiomes) et celles e qui placaient laccent sur les transformations lin aires et les vecteurs en faisant de e gros efforts pour r unier lenseignement de la g om trie avec celui des structures e e e alg briques. e
114
G. HOWSON
Dune mani` re g n rale, laccent etait mis presque enti` rement sur la g om trie e e e e e e qui devait etre apprise aux etudiants des etablissements d lite : gymnases, coll` ges, e e ` lyc es, . Cela cr a des probl` mes consid rables lorsque, a la n de ces d cennies, e e e e e les syst` mes scolaires devinrent moins s lectifs, avec pour but les math matiques e e e ` pour tous. Mais m me a lint rieur de ces secteurs d lite, avec des enseignants e e e math matiquement bien entran s, surgirent de s rieux probl` mes de formation continue. e e e e Non seulement les enseignants devaient apprendre les math matiques nouvellement e requises, mais il devenait aussi essentiel quils sachent appr cier les innovations dans e un cadre math matique plus large, an de saisir les buts vis s par les innovateurs pour e e ` se convaincre quils etaient a la fois d sirables et accessibles (du moins lorsque e ` cette afrmation etait justi e !). Cest donc par rapport aux lecons a tirer dans e lart du d veloppement curriculaire que l tude de ces d cennies apportera ses plus e e e grandes satisfactions. Cependant, lesprit dinvention, lenthousiasme et lambition ` math matique (quoique excessive) des innovateurs devraient servir daiguillon a ceux e qui sont m contents de lenseignement math matique qui est dispens dans les ecoles e e e aujourdhui.
GEOMETRY : 195070 by Geoffrey HOWSON
INTRODUCTION The rst thoughts of anyone studying the history of the teaching of geometry in the years 195070 must be of the degree of confusion concerning aims, the way in which these were implemented and also the wide range of outcomes. It was, indeed, a period of fertile instability 1 ). Everything seemed to be on the move but there was no consensus concerning the direction in which moves should be made. For that reason, it would be impossible in the space allowed to discuss all those initiatives of which I am aware and my knowledge is, of course, limited. What I shall attempt to do is to look briey at four key questions :
What caused the revolution in geometry teaching ? What was attempted ? What was achieved ? What was learned ? (Or, perhaps, What lessons might have been learned,
but often were not ?)
1 ) Freudenthal, in a very important paper [1963], pointed out that although ICMI had chosen the teaching of geometry as one of its themes for study in the years 195458, this aroused little interest (a report of the outcomes is to be found in LEnseign. Math. (2) 5 (1959)), yet once ICMI had chosen algebra and arithmetic as subjects for study in 195862, everyone suddenly became preoccupied with geometry. Freudenthals paper along with those referred to by Artin, Stone, and Lombardo-Radice [1963] stemmed from an ICMI seminar held in 1961 in Bologna on the theme teaching geometry at the secondary level. Other papers from that meeting can be found in LEnseign. Math. (2) 9 (1963). Freudenthals own attempts to develop school geometry courses at IOWO and what is now the Freudenthal Institute in Utrecht came after 1970.
116
G. HOWSON
THE
BACKGROUND
Many of the reasons for the reform movement in geometry were shared with other topics. In particular, these included a wish to update the teaching of mathematics, and also to rewrite school curricula and textbooks so that they would not only reect twentieth century mathematics and the way in which it was used, but also ensure that there was not a gulf between school and university mathematics 2 ). However, the position of geometry was, in most countries, different from that of other subjects : the aims for geometry teaching were far from universally shared. Here, for example, it differed from algebra. No matter how much algebra was actually included in the school curriculum, there was considerable agreement (up to the late 1950s) on what school algebra might mean. After the introduction of letters to denote numbers or variables, came the construction of algebraic formulae, followed by the formation and solution of linear equations, then quadratics, then simultaneous linear, the properties of the roots of quadratic equations, of cubics, and so on. However, the aftermath of the early 20th century school reforms had left countries following very different patterns in their geometry teaching. Moreover, the position was further complicated by the existence of various types of schools/educational institutions and the fact that these taught, and valued, different forms of geometry. The academic schools still retained a Euclidean tradition, although Euclid itself had been largely swept aside. The sequence in which theorems were taught and the degree of axiomatic rigour varied from country to country, but the actual theorems Pythagoras, what is termed either Thales Theorem or the Strahlensatz, the circle theorems, and congruence and similarity properties were still to be found in most curricula. Where there were differences, this mainly concerned the position of coordinate geometry, the attention given to 3-D work and the use made of analytical methods, particularly when, in the Senior High School, conics were introduced. There was, though, one very signicant exception. Following World War II, and during the period of US occupation, schooling in Japan had been reorganised on US comprehensive lines. Moreover, the mathematics curriculum had been recast so that it sought to emphasise the uses of mathematics
2 ) The relationship between school and university mathematics was explored at a symposium held in Geneva in honour of Henri Fehr (see, e.g., [Behnke 1957; Freudenthal 1956; Maxwell 1956]). Concern regarding the lack of accuracy, honesty and clarity in school texts was also expressed (see, e.g., [Cockcroft 1962]) at a seminar organised by ICMI in Lausanne in 1961.
GEOMETRY : 195070
117
in solving social problems. What was laid down for Japanese schools (see [Nagasaki 1990]) was not an academic curriculum, but one which, in Europe, would have been associated more with technical schools or Realschulen. That is, the mensuration of plane and solid gures, their simple properties, the ability to read and interpret maps, leading to consideration of congruence and similarity, an introduction to coordinates and the use of projection (what elsewhere might have been seen as the beginnings of technical drawing or descriptive geometry). Academic, Euclidean-style geometry formed an optional unit in the Japanese Upper Secondary School, but students disliked it, [few] pupils learned geometry and this caused difculties in the teaching of mathematics in the colleges and universities [Wada et al. 1956, 167]. The Japanese experience raises interesting questions : What form should geometry take in an education for all ? How are the competing claims of academic and artisan geometry (i.e. surveying, making and reading plans and maps, technical drawing (perspective, elevations and projections)) to be reconciled ? Can one overcome the fact that Euclid-style geometry is found extremely difcult (and often uninteresting) by most students 3 ) ? Can one ignore the benets of an academic geometry course for those students capable of following one ? And what exactly are those benets ? Such questions were given little consideration at the famous Royaumont conference of 1959 [OEEC 1961]. There, attention appeared to be almost entirely directed at students in selective schools and often it appeared to be assumed that the object of secondary school mathematics was solely to prepare students for its study at university. Indeed, it might be argued that, more than this, the hidden agenda was to determine which students were capable of such study. What is remembered most from that meeting is, of course, Dieudonn s e call that Euclid must go [Dieudonn 1961]. The call, however, was far e from novel and as we have seen there were countries in which Euclid had disappeared. (However, it should be noted that in Japan Euclid-style geometry re-entered the Lower Secondary Schools in 1958 and has remained there ever
3 ) Numerous quotes could be given to support this statement. A typical one is that by Tammadge [1987, 22] reporting on his experiences as an examiner of the top 20% or so of English students : Euclidean geometry had been watered down to the absurd. Only a small percentage of candidates attempted questions on this topic and they normally regurgitated the theorem and collapsed when it came to the rider [i.e. a request to prove a corollary to the theorem].
118
G. HOWSON
since. As two Japanese educators wrote, a curriculum based on the needs of society was changed to one based on the needs of mathematics [Kunimune & Nagasaki 1996].) Indeed, writing in 1911, D. E. Smith argued that
The efforts usually made to improve the spirit of Euclid are trivial [] but there is a possibility [] that a geometry will be developed that will be as serious as Euclids and as effective in the education of the thinking individual. If so, it seems probable that it will not be based upon the congruence of triangles, but upon certain postulates of motion []. It will be through some effort as this, rather than through the weakening of the EuclidLegendre style of geometry, that any improvement is likely to come. [Smith 1911]
Signicantly, Dieudonn himself recommended that up to the age of 14 e the teaching of geometry should be experimental and part of physics, so to speak. But the emphasis should not be on such articial playthings as triangles but on basic notions such as symmetries, translations, composition of transformations, etc.. An argument that in retrospect does not look terribly revolutionary. However, from age 15 he proposed the introduction of the axiomatic method :
The axioms should be developed from the algebraic and geometric point of view, i.e. any notion should be given with both kinds of interpretation. [] the emphasis should be on the linear transformations, their various types and the groups they form. Matrices and determinants of order 2 appear [] in a natural way in this development. [Dieudonn 1961] e
Important lessons were learned by those who attempted to follow this path, for neither Dieudonn nor anyone else had experience of how to communicate e axiomatics to adolescents and their teachers.
WHAT
HAPPENED ?
The call to reform geometry teaching was one that found a ready response. For a variety of reasons there was dissatisfaction with what was currently being offered. We shall here try to group the responses under a variety of headings. Again, it must be emphasised that all the initiatives were designed initially for students in the top 25% or so of the ability range. APPROACHES
IN THE SPIRIT OF
DIEUDONNE
OR IN WHAT WAS THOUGHT TO
BE THE SPIRIT OF
DIEUDONNE
Here Papys work must be mentioned. He believed that the most fundamental and central topic of the secondary school programme is, without
GEOMETRY : 195070
119
doubt, vector spaces and designed an approach which cleverly combined the development of the real number system and the afne plane 4 ). Many alternative approaches to Papys, often written with different age groups of readers in mind, were provided by, for example, Choquet (in an outstanding book LEnseignement de la G om trie 5 ), which was greatly to inuence 6 ) the e e school mathematics course in France), Glaymann [1969], and Levi [1971]. The emphasis on linear algebra 7 ) drew strong criticism from Freudenthal who argued that the geometry to which it lends itself is restricted and by no means the type to interest students. Heaven help the child brought up on Dieudonn s approach who seeks to prove that the plane can be tessellated e with regular hexagons but not with regular pentagons ! (See also [Freudenthal 1963] which criticises some of the schemes proposed for teaching geometry in the light of the work of the van Hieles, which at that time was little known
4 ) Papys books (Math matique moderne, Didier, 1963 on) aroused much interest and some e appeared also, for example, in German, Italian and English translations. They were distinguished by a most imaginative use of colour and diagrams. A rationale and synopsis of their contents can be found in [Papy 1966a; 1966b], and a summary of Papys approach to the number systems and geometry in [Grifths & Howson 1974, 282286]. In these it is described how, for example, 1213 year-olds reached the denition of a vector via the path : parallels, equipollency [(A B) and (C D) are equipollent if and only if AB is parallel to CD and AC to BD], equivalence classes, translation, vector. This method was adopted elsewhere and, for example, Leonid Brezhnev, not usually remembered for his contributions to mathematics education, in 1981 told the Central Committee of the Soviet Communist Party that, in future, there would be no more talk of equivalence classes : a vector would revert to being something having magnitude and direction. ([Keitel 1982] gives references to a report of the speech in Pravda, but no details of its contents other than that they spelled the end of the Kolmogorov reforms. I am relying on my memory of reading a translation of the speech when visiting Moscow in 1982.) It should be noted that rather than beginning with a list of axioms, Papy adopted what he described as the progressive axiomatic approach, namely : State clearly that which is accepted : do not say everything at one time; state certain accepted things, little by little. This method has some pedagogical benets, and at rst glance it does tend to show how an axiom system evolves. This latter benet is largely illusory, however, since only the author is aware of why he has selected a particular statement as an axiom (and the need for new axioms is not long in coming e.g. in proving that equipollency is transitive). SMSG and SSMCIS (see below) both made all their axioms explicit from the start and these were largely statements beyond dispute. Choquets axioms were very much a result of mathematical hindsight and at the time were called parachute axioms, i.e. they attack problems from the rear and were not framed because they were natural assumptions. 5 ) The title of the English translation, Geometry in a Modern Setting, gives a better indication of the actual nature of the work. 6 ) It should be noted that many articles were written, often by distinguished mathematicians on what and how geometry should be taught in schools. LEnseignement Math matique and e Educational Studies in Mathematics provide numerous examples of these. The vast majority, however, were never translated into classroom terms and use. Freudenthal [1963] compared this to producing theorems without proofs. 7 ) Although the natural link was between the new geometry and algebra, there were also attempts to link a new approach to geometry teaching with an introduction to analysis (see, e.g. [Delessert 1962]).

120
G. HOWSON
outside the Netherlands.) Other criticisms were to come from Artin [1963] who, in a very cogent article, reviewed what he referred to as three extreme views on the teaching of geometry 8 ). As he pointed out, the approach using linear algebra must, at some time, call on some subtle reasoning to deal with the introduction of the notion of angle 9 ). Strong criticisms of the new French curriculum 10 ) for the junior high school years (see [Adda 1981]), and of the inuences of Choquet and Dieudonn , were also to come from, amongst others, Thom [1970 ; 1973], e who argued for the retention of Euclid on a variety of grounds 11 ), and Leray [1966] 12 ). Perhaps, here, one should include an example to illustrate just how far the French reforms went (beyond anything envisaged by Dieudonn ) and e what was inicted upon both students and teachers. The following denition and theorem 13 ) were intended for the classe de quatri` me, i.e. for 14 year-olds. e
0 , lapplication

est bijective. 2 o La famille de toutes les bijections ainsi d nies poss` de la propri t : e e ee Pour deux bijections quelconques et de cette famille, il existe un e couple (a b) de nombres r els, tel que a 0 , et pour tout el ment M de : e
8 ) Essentially, teach Euclid and pretend it is rigorous, teach Hilbert (or a later geometrical axiom system), or use a basically algebraic axiom system. 9 ) A further, very interesting paper to appear at this time was by Marshall Stone [1963]. In it, Stone, takes as axiomatic that students between the ages of 15 or 16 and 18 should meet an axiomatic treatment of the Euclidean plane and space. He then goes on to lay down certain didactical rules that should characterise such a presentation. Inter alia, he refers to the work of Birkhoff and Beatley, Choquet, and the SMSG which are mentioned briey in this paper. Stone was, I believe, very much concerned as a consultant to the SSMCIS Project, yet (see below) that project did not develop an axiomatic presentation of the Euclidean plane but chose to demonstrate only how an axiom system worked. By this stage, the reader of the present paper should have noticed the unprecedented (and subsequently unmatched) interest that leading mathematicians were at that time displaying in school curricula. 10 ) Earlier reforms of the curriculum for the higher classes in France had proved less contentious. These introduced the various types of linear transformation and demonstrated their group properties. 11 ) Thom argued that although algebra might be stronger in syntax, Euclid-style geometry was stronger in meaning. 12 13
) A recent critical appraisal of the French reforms can be found in [CREM 2000]. ` e ) Reprinted in Les maths modernes a l cole in Sp cial Bourbaki, Pour la Science e
(f vrier 2000), p. 83. e
(M)
(M)
(M)
(M)
Etant donn e une droite gradu e ( ) . e e 1 o Pour tout couple de r els (a b ) tel que a e e sur R d nie pour tout el ment M de par : e

de
GEOMETRY : 195070
121
On appelle alors graduation de toute bijection de cette famille, et le nombre (M) est appel abscisse de M dans la graduation . 14 ) 15 ) e
No wonder that Leray should conclude a report to the French Academy of Sciences in 1972 by writing 16 ) :
Loption ensembliste de la d nition de la g om trie est une dangereuse utopie. e e e [] Les termes scientiques que nous avons du employer pour lanalyser montrent combien cette r forme m connat les aptitudes et besoins intellectuels e e e des adolescents qui sont el` ves des [] lyc es. La r forme en cours met e e gravement en danger lavenir economique, technique et scientique du Pays. 17 ) (quoted in [Adda 1981])
In other countries, vector geometry was also stressed (e.g. by Hope 18 ) in England) but here the emphasis was not so much on teaching the concept of a linear space, but frequently on using vector algebra to prove those theorems about triangles that Dieudonn so despised. e
is bijective. 2 o The family of all bijections thus dened has the following property : and from this family, there exists a pair (a b) For any two bijections of real numbers, with a 0, such that for every element M in :

(M)
(M)
We then call any bijection in this family a graduation of , and the number (M) is called the abscissa of M in the graduation . 15 ) An interesting, but not surprising, feature of the French and other reforms was the extra linguistic demands they made upon students. This, as several researchers demonstrated, led to students mathematical success being linked even more strongly to their home and social background.
16 ) Leray had expressed his concerns about some of the modern mathematics in a paper published by LEnseignement Math matique [Leray 1966]. e 17 ) The set-theoretic option in the denition of geometry is a dangerous utopia. [] The scientic terms which we have been forced to use in order to analyse it show how much this reform misappreciates the intellectual aptitudes and needs of the adolescents who attend our [] high schools. The reform in progress seriously endangers the economic, technical, and scientic future of the Nation. 18 ) In the early 1960s, Cyril Hope led an interesting project, the Midlands Mathematics Experiment, which, from the beginning, sought to bring new mathematics to students from a wide range of abilities. Its books had very many good points, both mathematically and pedagogically, but the project was inadequately nanced and it was overshadowed by the immensely more inuential SMP.
(M)
(M)
1 o For every pair of real numbers (a b ) such that a onto R dened for every element M in by :
14
) Given a line with a graduated scale ( ).
0, the map
from
122 APPROACHES
G. HOWSON
THROUGH TRANSFORMATION GEOMETRY
The ideas foreshadowed by D. E. Smith had gradually acquired greater currency in the 1950s and early 1960s. By that time the approach in Polish schools was already laying emphasis on the isometries and similarity transformations 19 ). Elsewhere authors such as Bachmann [1959], Yaglom [1962] and Jeger [1966] (again, writing for very different types of readership) were demonstrating the possibilities of an approach to geometry based upon linear transformations 20 ). An important example of this approach was that of the School Mathematics Project in England. Strangely, and by an extremely circuitous route, this came to resemble the pre-age-15 course described by Dieudonn (although devised e for somewhat older students). Originally, the course was intended as a two-year one to cover the age-range 1416. By then students would already have met some Euclidean-style geometry including Pythagoras. The rst year course comprised an experimental study of the isometries, enlargement and shearing : in Year 2 came the description of transformations in matrix terms, their composition (geometrically and algebraically) and the group of isometries. Those who attempted Additional mathematics, an examination for highattaining 16 year-olds, were offered the classication of frieze and plane patterns. Ideas of linearity were further developed in the 1618 course. Various attempts (e.g. [Maxwell 1975]) were made to provide an axiomatic basis and development for this approach to geometry teaching, but none took hold in the schools. NEW
AXIOMATIC APPROACHES TO PLANE GEOMETRY
In 1932 G. D. Birkhoff produced an axiom system which assumed the fundamental properties of the real numbers, basically permitting the use of a ruler and protractor in a natural way. As a result, using only four axioms, he was able to reach interesting theorems very quickly. Later, he and Beatley [1940] incorporated these axioms into a school text. A further attempt to
19 ) This approach was developed in a text book series by S. Kulczycki and S. Straszewicz (see [Ehrenfeucht 1978]). 20 ) Despite the pedagogical principle that one only fully understands a denition when one meets something that does not satisfy it, I can recall no examples of non-linear transformations being exhibited except towards the end of secondary education. Then, for example, SMPs syllabus for the very high-attaining mathematics students included inversion [and] simple conformal transformations to be studied amongst geometrical properties in the complex plane. (Conformal was dropped from the syllabus within a few years.)
GEOMETRY : 195070
123
proceed on these lines was made by the US School Mathematics Study Group (SMSG) on its establishment in 1958. A second US project, the Secondary School Mathematics Curriculum Improvement Study (SSMCIS) opted for a less ambitious axiom system 21 ) which, although not leading to the results of traditional Euclidean geometry, still gave students an inkling of axiomatic reasoning. Simple logical deductions were developed from these axioms and both geometric and non-geometric models of the axioms were exhibited. In many countries, however, a completely axiomatic approach was rejected (see, e.g., [Freudenthal 1971 ; Grifths & Howson 1974 ; Halmos & Varga 1978]) in favour of what was termed local organisation small packets of deductive geometry centred on, for example, the circle theorems. EXTENDING
THE BOUNDS OF SCHOOL GEOMETRY
A marked feature of geometry teaching in the 50s and 60s was the range of geometrical topics introduced into the school curriculum. The SMP, for example, included in its original draft texts (but not in the published versions) work on non-Euclidean geometries (including the models of Beltrami and Poincar ) and also some work on nite geometries 22 ). Its later texts, SMP e Books 15 23 ), for the age range 1116, contained some topology (e.g., Eulers formula), graph theory (including dual graphs and applications to operational research), the earth as a sphere (small and great circles, etc.), as well as artisan type work on surveying, perspective and technical drawing of threedimensional objects. Elsewhere, there were similar attempts to introduce new topics, although the approach was often very much more pure and abstract. For example, the US Contemporary School Mathematics Project [Kaufman 1971, 284] proposed that Grade 10 geometry should include afne geometry in 2- and 3-space with considerable emphasis on characterizing the nite afne planes, together with projective planes, particularly nite projective planes. The air of unrealistic optimism (generally based upon out-of-classroom discussions, or one-off demonstration classes 24 ) by enthusiasts) typical of the time is
21 ) The axiom systems of Birkhoff and SMSG are reprinted as appendices in [Tuller 1967], that for SSMCIS can be found in [Grifths & Howson 1974]. 22 23
) This was, however, non-examinable, enrichment work. ) These books, published by Cambridge University Press from 1965 onwards, aroused
considerable international interest and were translated, in whole or parts, into many languages, including Arabic, Chinese, French, Italian, and Spanish, as well as appearing in versions adapted for use in other anglophone countries.
24
) I am told that, under the aegis of the Institut National de la Recherche P dagogique, there e
124
G. HOWSON
exemplied by Kaufmans comments on these plans : This will be an easy task, as the usual correspondence between afne and projective planes will be established. Also duality will be stressed : thus, this should prove an extremely interesting and meaningful topic [Kaufman 1971, 284].
TOPSY-TURVYDOM A remarkable trait of many reformers was their rush to try something different, usually ignoring on the way the students existing knowledge and the type of mathematics he or she would meet outside the classroom 25 ). Piagets assertions concerning geometrical intuitions (see, e.g. [Piaget 1973, 83]) were often used to justify the early introduction of topology (as this rarely amounted to much it did neither great harm nor good). More signicantly, the inuence of Bourbaki (whose work greatly affected Piaget, particularly in what he referred to as the three, Bourbaki-revealed, mother structures of algebra, order, and topology) presumably led to the afne plane being frequently studied prior to the metric plane (see, e.g., [Kaufman 1971 ; Laborde 1998]). This was to cause considerably more problems. The arguments advanced for teaching afne before metric, and those against, can be found in [UNESCO 1973, 3031]. In the same chapter one nds what is to me an almost unbelievable statement :
In the afne approach students are made more and more aware that in order to do geometry one has only to remember the axioms and denitions pertaining to real numbers and those which give the structure of the afne vector plane.
were some classroom trials of the French materials but in non-representative high-attaining schools with enthusiastic and committed teachers. In England the SMP initially tested draft materials in eight or so schools most of which supplied authors to the project, and then second drafts were tested in a much wider, and more representative, range of schools before publication of a nal version. Even that did not resolve all the problems, for successful chapters tended to become more polished through the various versions, whereas approaches to difcult topics were often changed signicantly, even in some cases up to a largely untried nal version. The dangers of extrapolating from what committed, enthusiastic, well-informed teachers could achieve in the classroom to what would happen in the classrooms of typical teachers were very great and frequently ignored.
25 ) Such moves were much criticised by Freudenthal and Thom. Freudenthal [1963] was critical of what he called didactical inversion or anti-didactical inversion, e.g. the use of parachute axioms. Thom [1973] invoked Haeckels Law of recapitulation ontogenesis recapitulates phylogenesis to argue against abandoning historical approaches.
GEOMETRY : 195070
125
I write unbelievable, not just because it would seem to me to embody both an extremely limited view of geometry and what to do geometry might mean, but also because it was written, without any qualication, by a mathematics educator held in high regard. The tendency to break away from old patterns is further exemplied by Halmos and Vargas account [1978] of reforms in Hungary. There, it was argued, teaching should now proceed from the general to the particular. That is, the old-fashioned sequence of
3 FE4&C" 4B D 1 1 &$ 8 A4@ ) 9 6 754&&3 2)0( '&%$ # ! 1 1 "
should be replaced by
W%FBF%$ U2T&1S%FB5% V P 1 ) H 6 Q P 1 8 H 1 RB4G41 #&&I&G%FB
I nd it hard to see how such an approach could possibly succeed. Yet Varga 26 ) was certainly not the well-known mathematics educator whom Freudenthal once described as having learned his psychology from Bourbaki and his mathematics from Piaget !
WHAT
EFFECT DID THESE INNOVATIONS HAVE ?
In retrospect, it would be easy to dismiss the period 195070 as one of wasted effort and missed opportunities. So little of what was attempted still survives in schools today and geometry appears to many to be in a worse state than ever.
26 ) Varga and the anonymous author of the quotation on to do geometry were educators and teacher trainers of great merit who had enormous national and international inuence in their time and whose inuence is still felt through the work of their students. They should not be judged on these two quotations. Nevertheless the quotations have an important role in this paper in indicating the inuences of Bourbaki and his followers. The brilliance of the Bourbaki exposition, which, as Blakers (see below) indicated, encouraged mathematicians to see the mathematics they had acquired in a new and restructured manner, temporarily led some educators to imagine that it might also lead to a novel and more successful pedagogical approach. Only later was it accepted that Bourbakis approach was pedagogically and, it could be argued, mathematically sterile.
126
G. HOWSON
Yet there were some successes 27 ), e.g., the earlier introduction of coordinate geometry. Moreover, I believe the original SMP Books T/T4 28 ) had considerable merits in developing students spatial understanding, presenting modern concepts in an appropriate and achievable form, linking algebra and geometry, and laying the foundations for future mathematical learning. However, these successes were largely swept aside by changes within educational systems over which mathematicians and mathematics educators had no control and for which they were unable, or lacked the foresight, to prepare. All countries witnessed a great widening of opportunities in those two decades. In many, the compulsory age for school attendance was raised, but in all the proportion of the age cohort remaining in school beyond that age rose sharply. Several countries, particularly in Europe had to cope with the coming of comprehensive schools. This usually had the consequence that not only were the aims for high-attainers diluted, but the mathematics teaching specialists who previously had congregated in the selective schools were now more widely and thinly distributed. It is signicant that in England the original SMP Books 15 continued to be used (from the early 1980s in a revised, but essentially unchanged form) in some of the remaining independent (non-state) and selective schools until the advent of the National Curriculum in the early 1990s. However, in the vast majority of state schools the lower secondary school geometry curriculum became so diluted that by the mid-1980s matrices and the geometry/algebra links had entirely vanished. As in other countries, the transformation geometry that was by now commonly taught lacked welldened aims. Other signicant societal inuences were linked to economic changes for example, the way in which countries such as the Netherlands changed from agricultural to essentially industrial societies and the effect these had on the role of mathematics and mathematicians within society. Euclidean geometry no longer seemed so important, and there was not a steady stream of well-qualied mathematics graduates returning to schools to teach the subject now they were offered a much wider range of career opportunities.
27 ) Dening success is far from easy : does it mean sold a lot of copies or mathematically and pedagogically sound and appropriate for the target age group ? Certainly I should class the series of texts published in the 1970s by Lombardo-Radice (an Italian mathematician greatly interested in improving school mathematics, see, e.g. [Lombardo-Radice 1963]) and Lina Mancini Proia (a pioneering schoolteacher) as a success using the second criterion but I suspect it was not a great nancial success for the publishers. Another inuential Italian schoolteacher contemporary who wrote much on geometry was Emma Castelnuovo (see, e.g. [Castelnuovo 1966]), but whose work is perhaps better represented by, say, [Castelnuovo & Barra 1976]. 28 ) Published by the Cambridge University Press, 196465. Here I should declare a personal interest and possible bias, for I edited Books T and T4 and also Books 13 of the later series.
GEOMETRY : 195070
127
One result of these pressures was that geometrys place in the upper secondary school became seriously weakened. Statistics and probability, and other utilitarian topics, began to take preference over Menelaus and Ceva and a synthetic approach to conic sections. Moreover, rst-year university courses no longer always included a unit on geometry. What university mathematicians came to deplore was not the students absence of geometrical knowledge concerning triangles (as Dieudonn made clear), but that of proof, which it e was assumed came from the study of Euclidean geometry. For certainly, local organisation promised more in this direction than it ever achieved. (Normally because the packets were so small that the work on them had little effect on examination performance, with the result that teachers and students often neglected them.)
WHAT
LESSONS WERE TO BE LEARNED ?
Perhaps the most obvious lesson to be learned from the 50s and 60s was just how difcult it is to bring about serious curriculum changes. First there is the problem of designing sensible new curricula. Then, one is faced with the enormous task of explaining the new curriculum to teachers 29 ). This means, not only describing the new content, suggesting how it might best be taught and examined, and providing pupils texts, but also conveying the purpose of the changes, and convincing teachers that new goals are attainable and will prove of benet to their students. Certainly, the role played by university mathematicians came to be greatly questioned as the two decades came to an end. Did they see school mathematics as simply the rst step in an assembly line that would eventually turn out a mathematics graduate ? Had they realised the problems of translating abstract concepts into school terms and of providing motivation for their study ? For one fact became increasingly apparent : the idea that explaining mathematics clearly and logically would automatically yield success was ill-founded. One can only agree with Rosenbloom who wrote :
29 ) Only Alexander Wittenberg of the major critics of modern mathematics seemed fully to spell out the importance of the role of teachers : The most crucial single factor for sound teaching of mathematics is and remains the teacher. [Wittenberg 1965, 307] This paper, representative of his other writings, shows a great grasp not only of the mathematical problems and those of the learner, but also of effecting change within educational systems. Regrettably he died in 1965 and his warnings went largely unheeded.
128
G. HOWSON
During the last years a number of expositions of vectors in mathematics have appeared. [] We nd many of these [] unsatisfactory. Often the authors give no natural and non-trivial problems which lead the student to feel a need for vector concepts. A number of these presentations seem overly abstract. Some approaches overemphasise the afne aspects, which lend themselves to heavily algebraic treatment and neglect metric geometry. There is often a neglect of geometric intuition. [Rosenbloom 1969]
Here, I should like to draw specic attention to one of Rosenblooms points that might otherwise pass unnoticed : the manner in which many of the innovations were not accompanied by a bank of problems that could not only be given to students, but would also catch their attention and interest. Until we arrive at that ideal world, there will also be the need for a variety of suitable, sensible problems that can be set in tests and examinations. Several topics which had sound mathematical and pedagogic merits suffered in the 60s simply because they were brought into ill-repute by examiners lacking in vision and thought. Even when an approach combined both mathematical and psychological know-how this was no guarantee of success. I recall a textbook series, produced by a well-known cognitive psychologist who had a good command of mathematics, that failed simply because it was so dull : there was no air, nothing that would grasp the interest and attention of pupils and little to provide motivation. Yet, to ignore the mathematicians was to run the risk of providing texts that lacked obvious mathematical purpose, or even correctness (and as a reviewer of books I saw a number of horrors). It was clear that the skills and knowledge of mathematicians, psychologists and teachers who have day-to-day contact with students were all required. Moreover, their work had to be guided by knowledge of the way that the educational system in which they worked was moving and by the goals that society the government, parents and employers held for it. However, even the best designed curricula cannot be implemented overnight and the problems of actually effecting successful change in the classroom were greatly underestimated. A typical reaction was described by Blakers concerning Australia :
Many mathematics teachers gained new insights into the subject and hoped [often incorrectly] that these [] could be shared with their students 30 ). Another group of teachers was hostile to the changes, sometimes out of conservatism and sometimes out of genuine concern for their inappropriateness ; but the
30 ) Although Blakers is here writing of schoolteachers, his remarks must also surely hold good for many of the university mathematicians involved in the reforms.
GEOMETRY : 195070
129
majority [of teachers] were rather indifferent, being quite willing to teach different content and approaches [] provided that someone else would [] tell them how to go about it. [Blakers 1978, 153]
The words of Beeby, the New Zealand educator, describe the results of such a situation :
There is one thing that distinguishes teaching from all other professions, except perhaps the Church no change in practice, no change in the curriculum has any meaning unless the teacher understands it and accepts it. If a young doctor gives an injection under instruction, or if an architect as a member of a team designs a roof truss, the efciency of the injection or the strength of the truss does not depend upon his faith in the formula he has used. With the teacher it does. If he does not understand the new method, or if he refuses to accept it other than supercially, instructions are to no avail. At the best he will go on doing in effect what he has always done, and at the worst he will produce some travesty of modern teaching. [Beeby 1970, 46]
These are hard lessons to have to digest 31 ). Moreover the problem of recruiting sufcient high-quality teachers has increased everywhere and will add to the difculty of effecting curriculum change. Yet improvements must be made 32 ). It is important then that this is done bearing in mind all the constraints and that we do not fall into the trap of unthinking and unrealistic optimism that blighted so many of the innovations of the period 195070.
REFERENCES ADDA, J. Etude de cas : la r forme des math matiques modernes. Bulletin de e e lAssociation francophone de l ducation compar e (1981), 117149. e e ARTIN, E. Les points de vue extr mes sur lenseignement de la g om trie. LEnseign. e e e Math. (2) 9 (1963), 14.
31 ) Hard facts concerning the difculties of major curricular innovation are most effectively spelled out in [Dalin 1978]. The author had directed a major project of the Centre for Educational Research and Innovation (CERI an offshoot of OECD) which considered the effectiveness of curriculum reform in a number of countries. Several volumes of case studies were published by OECD, but it is Dalins summary that is recommended reading to all interested and concerned with curriculum development. 32 ) Data from the Third International Mathematics and Science Study (TIMSS) relating to the pure geometrical knowledge of 18+ specialist mathematics students in their last year of secondary school make dismal reading. For example, an open response item asking students to prove a triangle was isosceles (given angle facts relating to two of its altitudes) was answered correctly by fewer than 10% of US specialist students; the corresponding percentages for two groups of high-scoring students, from Greece and France, were 65 and 53 respectively. (Specialist students represented approximately 14%, 10% and 20% respectively of the three countries age cohorts.) However, before one reads too much into these results and the students abilities to think logically, it should be noted that the percentages of students able to identify which of four statements could be logically deduced from the fact that (to paraphrase) 21 letters had been placed in 4 pigeon holes were : US, 70.0; France, 61.8 and Greece, 33.8.
130
G. HOWSON
BACHMANN, F. Aufbau der Geometrie aus dem Spiegelungsbegriff. Springer, Berlin, 1959. BEEBY, C. E. Curriculum planning. In : A. G. Howson (ed.), Designing a New Curriculum, 3952. Heinemann Educational, London, 1970. BEHNKE, H. La tension entre lenseignement secondaire et lenseignement universitaire en Allemagne. LEnseign. Math. (2) 3 (1957), 237250. BIRKHOFF, G. D. A set of postulates for plane geometry. Annals of Maths 33 (1932), 329345. BIRKHOFF, G. D. and R. BEATLEY. Basic Geometry. Scott Foresman, 1940. BLAKERS, A. L. Change in mathematics education since the late 1950s Ideas and realisation : Australia. Educational Studies in Mathematics 9 (1978), 147158. CASTELNUOVO, E. Un enseignement moderne des math matiques dans le 1er cycle e secondaire. LEnseign. Math. (2) 12 (1966), 195199. ` CASTELNUOVO, E. and M. BARRA. Matematica nella realta. Boringhieri, Torino, 1976. COCKCROFT, W. H. Some notes on British calculus text books. LEnseign. Math. (2) 8 (1962), 163170. CREM (COMMISSION DE REFLEXION SUR LENSEIGNEMENT DES MATHEMATIQUES). [2000] Rapport d tape sur la g om trie et son enseignement. In : Rapport au e e e ministre de l ducation nationale sur lenseignement des sciences math matiques, e e sous la dir. de J.-P. Kahane. O. Jacob/CNDP, Paris, 2002. DALIN, P. Limits to Educational Change. Macmillan, 1978. ` DELESSERT, A. Lenseignement de la g om trie comme pr paration a lanalyse. e e e LEnseign. Math. (2) 8 (1962), 136149. DIEUDONNE, J. New thinking in school mathematics. In : [OEEC 1961], 3145. EHRENFEUCHT, A. Change in mathematics education since the late 1950s Ideas and realisation : Poland. Educational Studies in Mathematics 9 (1978), 283295. FREUDENTHAL, H. Relations entre lenseignement secondaire et lenseignement universitaire en Hollande. LEnseign. Math. (2) 2 (1956), 238249. Enseignement des math matiques modernes ou enseignement moderne des e math matiques ? LEnseign. Math. (2) 9 (1963), 2844. e Geometry between the devil and the deep sea. Educational Studies in Mathematics 3 (1971), 413435. GLAYMANN, M. Initiation to vector spaces. Educational Studies in Mathematics 2 (1969), 6979. GRIFFITHS, H. B. and A. G. HOWSON. Mathematics : Society and Curricula. Cambridge University Press, 1974. HALMOS, M. and T. VARGA. Change in mathematics education since the late 1950s Ideas and realisation : Hungary. Educational Studies in Mathematics 9 (1978), 225244. JEGER, M. Transformation Geometry. Allen and Unwin, London, 1966. (English transl. of Konstruktive Abbildungsgeometrie.) KAUFMAN, B. A. Background : The Comprehensive School Mathematics Program. Educational Studies in Mathematics 3 (1971), 281285. KEITEL, C. Mathematics education and educational research in the USA and USSR. J. of Curriculum Studies 14 (2) (1982), 109126.
GEOMETRY : 195070
131
KUNIMUNE, S. and E. NAGASAKI. Curriculum changes on lower secondary school mathematics of Japan Focused on geometry. In : M. Kindt et al. (eds.), Curriculum Changes in the Secondary School, 1317. Freudenthal Institute, Utrecht, 1996. LABORDE, C. Geometry behind the French national curricula in the last decades. In : C. Mammana and V. Villani (eds.), Perspectives on the Teaching of Geometry for the 21st Century, 214222. Kluwer, 1998. LERAY, J. Linitiation aux math matiques. LEnseign. Math. (2) 12 (1966), 235241. e LEVI, H. Geometric algebra for the high school program. Educational Studies in Mathematics 3 (1971), 490500. LOMBARDO-RADICE, L. Geometria e cultura in un liceo moderno. LEnseign. Math. (2) 9 (1963), 6475. MAXWELL, E. A. From secondary school to University. LEnseign. Math. (2) 2 (1956), 307313. Geometry by Transformations. Cambridge University Press, 1975. NAGASAKI, E. (ed.) A History of the School Mathematics Curriculum in Japan. NIER, Tokyo, 1990. OEEC (ORGANISATION FOR EUROPEAN ECONOMIC CO-OPERATION). New Thinking in School Mathematics. OEEC, Paris, 1961. PAPY, G. [1966a] La g om trie dans lenseignement moderne de la math matique. e e e LEnseign. Math. (2) 12 (1966), 225233. [1966b] La g om trie dans lenseignement moderne de la math matique. In : e e e UNESCO (1966), 283293. PIAGET, J. Comments on mathematical education. In : A. G. Howson (ed.), Developments in Mathematics Education, 7987. Cambridge University Press, 1973. ROSENBLOOM, P. Vectors and symmetry. Educational Studies in Mathematics 2 (1969), 405414. SMITH, D. E. The Teaching of Geometry. Boston and London, 1911. ` e STONE, M. H. Les choix daxiomes pour la g om trie a l cole. LEnseign. Math. (2) e e 9 (1963), 4555. TAMMADGE, A. R. A mathematics master in the 1960s. In : A. G. Howson (ed.), Challenges and Responses in Mathematics, 2133. Cambridge University Press, 1987. THOM, R. Les Math matiques Modernes : Une erreur p dagogique et philosophique ? e e In : LAge de la Science 3 (1970), 225236. Modern mathematics : does it exist ? In : A. G. Howson (ed.), Developments in Mathematics Education, 194209. Cambridge University Press, 1973. TULLER, A. A Modern Introduction to Geometries. Van Nostrand, 1967. UNESCO. New Trends in Mathematics Teaching, Vol. I. UNESCO, Paris, 1966. New Trends in Mathematics Teaching, Vol. III. UNESCO, Paris, 1973. WADA, Y. et al. Report on mathematical teaching in the educational system of school and undergraduate courses in Japan, 1956. (Reprinted in : [Nagasaki 1990], 115 182.) WITTENBERG, A. Priorities and responsibilities in the reform of mathematical education : an essay in educational metatheory. LEnseign. Math. (2) 11 (1965), 287308. YAGLOM, I. M. Geometric Transformations. (English transl.) Random House, 1962.
` GEOMETRIE PERIODE 2000 ET APRES Geometry Period 2000 and after by Colette LABORDE
The teaching of geometry lost importance in the second half of the 20 th century, but there is now a widespread agreement on the necessity of maintaining and renewing it. Four main issues for geometry learning and teaching in the 21st century are discussed : the development of spatial awareness ; the dual nature of geometry and the relationship between theoretic and intuitive geometrical knowledge ; the learning process to formal reasoning ; the use of Information and Communication Technology. THE DEVELOPMENT OF SPATIAL AWARENESS. Geometry is rooted in actions and motions in the material environment. However, spatial awareness does not arise spontaneously and school should contribute to its development. Evidence has been gathered showing that children principally develop controls and spatial knowledge of a small-size space (microspace, the space of objects that they can take in their hands and manipulate) and have inadequate representations of both mesospace (from 0.5 to 50 times the size of an individual) and macrospace (in which objects cannot be grasped by hand or eye, e.g. the space of the city). The teaching of geometry deals mainly with small objects and diagrams which are drawn on a sheet of paper. But it is also essential to deal with modelling relations in meso- and macrospace with concepts like direction and angle. Research has also demonstrated that elementary-school children are unable to identify geometric relations even on diagrams of microspace. They are attracted by special features of the diagrams which are not relevant from a geometric point of view : segments are not considered to be parallel if they are not of the same length. Students should therefore learn to detach the geometric invariants from their visual appearance. THE DUAL NATURE OF GEOMETRY : TWO REGISTERS. Geometry is based on the use of two registers, the register of diagrams and the register of language. Language allows one to describe geometric objects and relations using specic terminology, while diagrams in 2D geometry play an ambiguous role. On the one hand they refer to theoretic objects ; on the other hand they offer graphical-spatial properties, which can generate a perceptual activity from the individual. In the traditional way of teaching geometry the theoretic properties are assimilated to graphical ones, with the illusory aim to abstract from a diagram the properties of the theoretic object it represents. As
134
C. LABORDE
a consequence, pupils often draw the conclusion that one can construct a geometric diagram using only visual cues ; or deduce a property empirically by checking on the diagram. When pupils are asked to construct a diagram, the teacher would like them to work at the geometry level, using theoretical knowledge. In reality they very often reason at the graphic level and try to satisfy visual constraints only. Teaching the distinction between spatial graphic relations and theoretical geometric relations should certainly devolve on geometry classes, but pupils should also integrate the ability to move between theoretical objects and their spatial representations, as well as the ability to recognize geometric relations in a diagram and to imagine all possible diagrams attached to a geometric object. This is particularly time-consuming for 3D geometry and little emphasis has been laid on its teaching, except in some countries such as the Netherlands in which 3D and 2D geometry are taught in interaction. The use of ICT is also of great help to compensate the lack of familiarity with 3D geometry. REASONING IN GEOMETRY. As research evidence suggests, numerous conceptual difculties arise in the learning of formal proof. Students fail to distinguish between empirical and deductive arguments ; in general they rather turn to empirical arguments. Two approaches have been developed worldwide : one in which writing a proof is detached from the heuristic search leading to it : the text of a proof is viewed as a specic text based on the distinction between the epistemic value of a proposition (true, false, plausible) and its theoretic status (hypothesis, conclusion, theorem). What matters is only the theoretic status of propositions, since all propositions involved have the epistemic value true ; an integrative approach in which proof as a product is the result of a long process of exploring, conjecturing, and arguing. Long-term projects of such innovative teaching have been especially developed in Italy. The fact that a proof performs a diversity of functions is now accepted by the research community, as well as the necessity of developing this diversity in teaching. In particular, proof as a means of explanation of visual phenomena seems to play a critical role in the coming era which will give more room to dynamic geometry environments offering enhanced visualization. THE USE OF ICT FOR TEACHING AND LEARNING GEOMETRY. Spatio-graphical and geometric aspects are very much interrelated in the new kind of diagrams provided by dynamic geometry computer environments. In such environments the diagrams result from a sequence of primitives which are expressed in geometric terms chosen by the user. When an element of such a diagram is dragged by means of the mouse, the diagram is modied but all geometric relations used in its construction are preserved. These articial realities could be compared to entities of the real world : they appear to react to the users manipulations by following the laws of geometry in the same way as material objects obey the laws of physics. A crucial feature of these realities is their quasi-independence from the user as soon as they have been created. When the user drags one of its elements, the diagram is modied according to the geometric way dened by its construction and not as the user wishes. Computer diagrams are external objects, whose behaviour requires the construction of an interpretation by the students. Geometry is a means, among others, of achieving this interpretation.
` GEOMETRIE PERIODE 2000 ET APRES par Colette LABORDE
1. UN
SUJET DE PREOCCUPATION DANS LE MONDE ENTIER
La n de lann e 2000 a clos une deuxi` me moiti de si` cle mouvement e e e e e e ` relativement a lenseignement de la g om trie, qui dune grande uniformit sur e e e le plan international avant les ann es 60, est pass a une forte h t rog n it a e e` ee e e e ` ` lheure actuelle [Hoyles et al. 2001]. Il est cependant loisible de se persuader a la lecture de l tude ICMI Perspectives on the Teaching of Geometry for the e 21st Century [Mammana & Villani 1998] 1 ) que des interrogations communes et des r gularit s transcendent cette grande variabilit de la place et du contenu e e e ` de la g om trie dans les curricula des diff rents pays du monde, a savoir : e e e le constat banal selon lequel la g om trie enseign e nest plus ce quelle e e e etait avant les ann es 60 ; e la n cessit de maintenir, voire de renforcer lenseignement de g om trie e e e e e ` ` de l cole el mentaire a la n de l cole secondaire (de 6 a 18 ou 19 ans) ; e e le souci de pr server les liens de lenseignement de la g om trie et de la e e e d monstration, tout en en r novant les modalit s ; e e e la n cessaire prise en compte de lapparition de logiciels de construction e g om trique et/ou de g om trie dynamique. e e e e Il nest pas anodin quen France, la commission de r exion sur e lenseignement des math matiques, pr sid e par Jean-Pierre Kahane, ait proe e e duit un rapport d tape sur la g om trie et son enseignement [CREM 2000] 2 ), e e e alors que simultan ment en Angleterre, pays de tradition scolaire fort diff rente, e e un groupe de travail r chissait sur lenseignement et lapprentissage de la e e
1 ) Voir en particulier le chapitre 7 : Changes and trends in geometry curricula. Voir aussi le compte rendu fait de cette etude par V. Villani dans LEnseign. Math. (2) 46 (2000), 411415. 2
) http://smf.emath.fr/Enseignement/CommissionKahane/
136
C. LABORDE
` g om trie a linitiative de la Royal Society et du Joint Mathematical Council e e et publiait en juillet 2001 les conclusions de sa r exion et ses recommandae tions 3 ). A la suite de l branlement qua subi lenseignement de la g om trie e e e ` un peu partout dans le monde a loccasion de la r forme des math matiques e e modernes, se sont faites jour, ici et l` , apr` s les ann es 80, des adaptations a e e diverses, quelquefois sans v ritable coh rence interne, qui ont conduit de facon e e ` ` internationale a sinterroger sur lenseignement de la g om trie et a soutenir e e son enseignement. Les raisons pour enseigner la g om trie sont aussi assez e e universellement partag es. Elles peuvent etre r sum es par les quelques points e e e suivants : le d veloppement pour tout individu de la matrise de lenvironnement e spatial ; les liens entre g om trie et vision ou intuition spatiales ; e e lapprentissage du raisonnement ; lutilit de connaissances g om triques dans la vie courante et limpore e e tance de la g om trie dans de nombreuses applications contemporaines e e (courbes de B zier, imagerie m dicale, infographie, images de synth` se, ) e e e et dans dautres sciences comme la physique, la chimie, la m canique ou e la biologie. Ces raisons ont une incidence certaine, tant sur les objectifs que peut poursuivre un curriculum de g om trie dans les ann es a venir que sur le e e e ` choix des contenus et des formes denseignement. Notre expos reprendra e certains des points list s ci-dessus et y ajoutera le r le possible jou par les e o e nouvelles technologies. Il prendra evidemment en compte les recherches des dix derni` res ann es sur lenseignement et lapprentissage de la g om trie e e e e dans diff rents pays du monde. e
2. LAPPRENTISSAGE
DE LA MATRISE DE LESPACE I
La g om trie, en tant que th orie, sest d velopp e sous des exigences e e e e e ` multiples, dont a un extr me celle de matriser lespace dans lequel vit e ` lhumanit , et a lautre de r soudre des probl` mes issus de la th orie m me. e e e e e Un des premiers objectifs de lenseignement de la g om trie est bien celui e e ` ` dapprendre a lenfant a saisir les situations spatiales dans toute leur complexit e ` [Rouche 2000]. Les connaissances spatiales sont celles qui permettent a
3 ) Teaching and Learning Geometry 1119, Policy document 15 (01 July 2001). The Royal Society, London.
` GEOMETRIE PERIODE 2000 ET APRES
137
lindividu dagir et de communiquer dans son environnement spatial : prendre un objet, se d placer en tenant compte des obstacles, rep rer la position e e relative dobjets, d crire et repr senter des sc` nes spatiales. Il est clair que la e e e g om trie est un outil de mod lisation puissant pour contr ler, agir, anticiper e e e o sur lespace puisque, comme dit plus haut, elle tire en partie son origine de ` e ` probl` mes spatiaux a r soudre. On doit a Freudenthal [1973] davoir insist e e sur cette fonction essentielle de la g om trie qui a pu etre oubli e lors de la e e e r forme des math matiques modernes. e e DES ` CONNAISSANCES SPATIALES A CONSTRUIRE
Mais si les racines famili` res (selon lexpression de Rouche [2000]) e de la g om trie se trouvent bien dans les actions sur lespace mat riel, le e e e d placement et le rep rage, il importe de reconnatre que ces connaissances e e spatiales ne sont pas construites de facon spontan e par les enfants et e sont souvent inexistantes chez eux, d` s que la taille de lespace devient e grande [Berthelot & Salin 1998, 7273]. Lespace qui nous entoure exige des connaissances et des contr les de nature spatiale fort diff rents suivant sa o e taille. Brousseau [1983] propose ainsi de distinguer trois espaces : le micro-espace, celui dont la taille permet de prendre des objets, les bouger ; le m so-espace, e celui dans lequel on vit, se d place, et que la vision peut appr hender (les e e objets mesurent entre 0,5 et 50 fois la taille de lindividu), par exemple celui de la salle de classe ; le macro-espace, celui dans lequel on ne peut atteindre les objets ni par le geste ni par la vue, par exemple celui de la ville. Si des e e ` el` ves de n d cole el mentaire ont appris a reconnatre sur une feuille de e papier que le dessin dun quadrilat` re est un carr en mesurant ses c t s, ils e e oe ` echouent largement a la t che de reproduire un rectangle de taille celle de la a base dun banc, se fondant toujours sur des mesures de longueur et ne sachant pas contr ler les angles [Berthelot & Salin 1998]. o ` La g om trie enseign e a en effet trop tendance a sappuyer sans contr le e e e o sur un rapport privil gi a lespace r serv au traitement de petits objets ou e e ` e e de trac s tenant sur une feuille de papier, sur l vidence perceptive : on voit e e bien que . Il importe donc d largir les moyens daction et de controle des e ` e e enfants sur lespace, en introduisant a l cole el mentaire des connaissances g om triques comme celle de vis e ou dangle. Les outils de mod lisation e e e e ` quapporte la g om trie seront alors non seulement susceptibles de contribuer a e e la matrise de lespace mais ils seront aussi la source de ph nom` nes qui posent e e question gr ce aux connaissances g om triques d j` disponibles et qui sinon a e e ea
138
C. LABORDE
passeraient inapercus. Lenseignement de la g om trie se doit dinstaurer une e e ` dialectique entre spatial et g om trique qui permette a lenfant de progresser e e sur les deux plans. Dautres recherches [Argaud 1998] montrent que m me les connaissances e e du micro-espace de la feuille de papier sont tr` s limit es chez les el` ves de e e e n d cole el mentaire. Argaud [1998, 290297], qui a travaill de facon tr` s e e e ` e e ` approfondie a l cole el mentaire, a pu constater a plusieurs reprises chez e ` ` des el` ves de 9 a 11 ans l chec important a des t ches de reconnaissance e a spatiale dun quadrilat` re, de deux segments parall` les, de deux segments e e perpendiculaires. Voici un exemple de r ponses obtenues dans une classe de e CM2 (1011 ans) : a) D estil un quadrilat` re (Fig. 1) ? e
FIGURE 1 e 14 el` ves sur 26 r pondent oui e
b) J est-il un quadrilat` re (Fig. 2) ? e
FIGURE 2 e e e 10 el` ves sur 13 r pondent oui. Les arguments sont : il a 4 cot s (8 el` ves), e e e m me sil nest pas ni, il a 4 c t s (1 el` ve), il a 4 ar tes (1 el` ve). e oe e
139
c) Les segments sont-ils parall` les (Fig. 3) ? e
FIGURE 3 e Seize el` ves sur vingt-quatre disent quils ne sont pas parall` les. Apr` s discussion en bin mes, e e o cinq paires afrment que les segments ne sont pas parall` les, avec des arguments comme : e ils ne sont pas en face, ils sont d cal s, ils ne sont pas de la m me longueur . e e e Cinq autres paires d clarent que les segments sont parall` les (avec des arguments comme : e e si on les prolonge, ils ne se touchent pas ). Deux paires ne se prononcent pas.
e Comme on peut le constater, la difcult r side pour les el` ves dans e e lidentication des propri t s spatiales pertinentes qui caract risent la propri t ee e ee e g om trique concern e. De la m me facon, les el` ves d cole moyenne (1115 e e e e e ans) ne reconnaissent pas ais ment un losange, voire un carr sils ne sont pas e e dans une position prototypique, ou un axe de sym trie sil nest pas vertical. e Or, si la g om trie contribue au d veloppement de connaissances spatiales, e e e elle sappuie egalement sur ces derni` res car le raisonnement g om trique est e e e fond sur une interaction entre le voir et le savoir [Parzysz 1988], ou encore e selon les termes de Fishbein [1993] entre gural et conceptuel. La g om trie e e est de nature duale et cest une de ses caract ristiques qui la rend si pr cieuse e e et qui motive son enseignement.
3. LA
NATURE DUALE DE LA GEOMETRIE : LE SENSIBLE ET LINTELLIGIBLE
On sait que la Gr` ce antique avait d j` d battu de la place de la g om trie e ea e e e e relevait-elle du sensible ou de lintelligible ? et que la d cision avait et e prise en faveur des fameuses id alit s. La nature duale de la g om trie a depuis e e e e e et maintes fois exprim e, comme le font encore Hilbert et Cohn-Vossen dans e la pr face de louvrage Anschauliche Geometrie : e
In der Mathematik wie in aller wissenschaftlichen Forschung treffen wir zweierlei Tendenzen an : die Tendenz zur Abstraktion sie sucht die logischen Gesichtspunkte aus dem vielf ltigen Material herauszuarbeiten und dieses in a
140
C. LABORDE
systematischen Zusammenhang zu bringen und die andere Tendenz, die der Anschaulichkeit, die vielmehr auf ein lebendiges Erfassen der Gegenst nde a und ihre inhaltlichen Beziehungen ausgeht. 4 ) [Hilbert & Cohn-Vossen 1932, v]
La g om trie, en tant que domaine des math matiques, met en jeu des e e e objets et des relations th oriques mais elle est issue dune mod lisation de e e lespace mat riel qui nous environne. Toucher, couper, avoir une forme e ronde, tre pr` s de, sont autant de locutions d signant des ph nom` nes e e e e e spatiaux que la g om trie exprime en termes de relations entre des objets e e d nis par la th orie : les droites sont s cantes , la droite est tangente au e e e cercle , Dans certains cas, le langage de lespace et celui de la g om trie e e ont recours aux m mes termes, lorsquil sagit de rendre compte de propri t s e ee ` courantes : etre parall` le, se couper a angle droit. Il importe cependant de e ` distinguer deux r f rents diff rents, si lon cherche a mieux comprendre les ee e e difcult s dapprentissage des el` ves en g om trie : dun c t , un domaine e e e oe th orique 5 ) issu dune mod lisation de lespace mais se d veloppant ensuite e e e ` ` en grande partie a partir de probl` mes internes a la th orie et ayant son e e propre mode de validation (d monstration), de lautre une r alit spatiale e e e li e a notre environnement que certains appellent g om trie exp rimentale e ` e e e [Chevallard 2001] ou naturelle [Houdement & Kuzniak 1999], dans laquelle la validation se fait par lexp rience sensible. e EXISTENCE EN GEOMETRIE DE DEUX REGISTRES SEMIOTIQUES
Les objets et relations de la g om trie sont ext rioris s soit dans un registre e e e e de type discursif, par des enonc s dans des langages plus ou moins formels, e soit dans un registre de type gural, par des dessins ou encore des entit s e ` que nous appellerons spatio-graphiques, retournant ainsi en quelque sorte a lorigine spatiale de la g om trie. Ces deux types dexpression des objets e e g om triques sollicitent une appr hension ainsi que des connaissances et des e e e contr les de nature diff rente de la part de lindividu. Les dessins appellent o e ` une appr hension globale en donnant a voir des relations spatiales, tandis que e les enonc s sollicitent une appr hension lin aire et analytique dun discours e e e
4 ) En math matiques, comme dans toute recherche scientique, on trouve deux types de e ` ` e tendances : la tendance a labstraction (elle cherche a d gager les aspects logiques de la diversit e ` du mat riel etudi et a corr ler celui-ci de facon syst matique et coh rente) et, dautre part, la e e e e e ` ` tendance a la conceptualisation intuitive, qui aspire plutot a une compr hension vivante des objets e en se concentrant sur le fondement concret de leurs rapports r ciproques. e 5 ) Certes, on peut distinguer des niveaux dafnement de la th orie au sein m me de la e e g om trie th orique. En particulier, nombre de d monstrations effectu es a l cole secondaire e e e e e ` e comportent des imperfections dont on sait quelles seraient susceptibles d tre elimin es dans une e e g om trie que Chevallard [2001] qualie de th orique math matique. e e e e
141
qui renvoie plut t aux objets th oriques g om triques. Les contr les mis en jeu o e e e o par lappr hension du dessin sont en un premier temps de type perceptif, tandis e que les contr les sur les enonc s se font par des connaissances g om triques. o e e e Cest bien dabord la reconnaissance visuelle du contact dune droite avec un ` cercle qui nous conduit a identier une tangente sur un dessin et cest en un deuxi` me temps que gr ce a nos connaissances g om triques, nous nous e a ` e e assurons du bien-fond g om trique de cette impression visuelle, par exemple e e e en v riant que la droite est bien perpendiculaire au rayon. e LA DISTINCTION ENTRE SPATIO-GRAPHIQUE ET GEOMETRIQUE
e Il est evident que de facon spontan e les el` ves de n d cole primaire e e et de d but de coll` ge ne peuvent distinguer le spatio-graphique du th orique e e e ` et que cest a lenseignement qu choit la mise en place de cette distinction, e qui est une entreprise de longue haleine. De nombreux constats de cette e confusion chez les el` ves ont pu etre faits. L` o` lenseignement attend a u e ` une mise en uvre dune propri t g om trique, les el` ves ont recours a ee e e ` un constat visuel, comme dans le cas de la construction dune tangente a e un cercle, issue dun point P ext rieur au cercle : les el` ves la tracent par e t tonnement sur le dessin en mettant en uvre un controle spatial, la bonne a ` e droite est obtenue a lil quand la r` gle touche le cercle. Les el` ves travaillent e spontan ment sur les propri t s spatiales du dessin dans un contrat graphique e ee ` alors que lenseignement cherche a mettre en place un contrat g om trique e e [Arsac 1993]. LE JEU ENTRE GEOMETRIQUE ET SPATIO-GRAPHIQUE
` Il semble donc que non seulement le th orique soit a construire par les e e el` ves, mais aussi les rapports entre des propri t s g om triques th oriques et ee e e e les repr sentations spatio-graphiques. Linterpr tation dun dessin comme celui e e dun objet g om trique, la prise en consid ration de ce dessin comme une e e e repr sentation parmi dautres de lobjet g om trique en question consiste en e e e ce que nous appelons la gure [Laborde & Capponi 1994]. En g om trie plane e e ` euclidienne, on a peine a imaginer quil soit difcile daller au-del` du dessin a en tant quentit mat rielle et de consid rer la (ou les) gure(s) attach e(s) e e e e au dessin, tant est fort le sentiment d vidence que lon voit la gure. Le e passage en g om trie non-euclidienne, par la d centration quil occasionne, e e e permet de mieux prendre conscience que lassociation entre le dessin et lobjet g om trique quil repr sente ne va pas de soi. En effet, la g om trie none e e e e euclidienne a recours aux dessins de la g om trie euclidienne qui ne sont alors e e
142
C. LABORDE
plus analogiques comme ils l taient en g om trie euclidienne : une droite nest e e e ainsi plus repr sent e par un trait rectiligne dans le mod` le de Poincar . Le e e e e r le symbolique du dessin est plus apparent, comme l crit Turc [1914] : o e
Ne pouvant construire les gures lobatschewskiennes, cest dune mani` re e symbolique que lon trace les gures planes sur un plan et que lon y dessine les gures de lespace. Cette repr sentation symbolique suft pour xer les e id es et suivre le raisonnement. e
Certes le sentiment d vidence au vu du dessin peut aussi etre obtenu, e mais au prix dune longue pratique. Thibault et Labarre [1996, 221] indiquent combien des etudiants canadiens dun cours de g om trie de premier cycle e e universitaire en math matiques et en education math matique ont des difcult s e e e ` a reconnatre une droite dans le dessin dune ligne courbe dans un mod` le e de g om trie hyperbolique et combien le travail dans plusieurs mod` les de e e e g om tries non-euclidiennes peut etre fructueux pour d passer ce type de e e e difcult . Arsac [1998] tire les m mes conclusions sur lint r t dun tel travail e e ee en formation des matres, qui permet de sensibiliser les form s aux nombreux e ` postulats implicites en cours dans la g om trie euclidienne enseign e et a la e e e prise de conscience que l vidence est toute relative. e e Or, il nous parat essentiel que les el` ves sachent utiliser un jeu dinteractions entre contr les spatiaux et contr les th oriques. En effet, la o o e compl mentarit des deux registres et des appr hensions et contr les joue un e e e o r le important dans la r solution de probl` mes en g om trie, comme lont o e e e e soulign diff rents travaux [Fishbein 1993 ; Mariotti 1995 ; Duval 1998]. Cest e e parce que lexpert sait mettre en place une interaction entre ces deux registres, quil avance dans la r solution dun probl` me de g om trie. Il sait distinguer e e e e les ph nom` nes spatiaux pertinents pour le probl` me, des propri t s spatiales e e e ee non int ressantes, il sait interpr ter en termes de g om trie ce quil voit, ou e e e e m me il sait gr ce a la g om trie voir plus dans le dessin que ce que lon peut e a ` e e y voir de facon imm diate, et enn il sait tirer des cons quences g om triques e e e e de ce quil constate sur le dessin. Rouche [2001, 51] consid` re que la pens e e e g om trique commencante met en jeu trois processus percevoir, concevoir e e ` et inf rer qui sint` grent dans lactivit . Mener a bien un raisonnement e e e en g om trie repose sur une interaction entre le registre discursif et celui des e e gures, entre lappr hension de propri t s spatiales et lusage de propri t s e ee ee th oriques. e Un des objectifs de lenseignement de la g om trie r side donc dans e e e lorganisation de situations dapprentissage de traitements des dessins en e g om trie, o` il sagit pour les el` ves de d passer ce que Duval [1998] e e u e appelle lappr hension perceptive imm diate qui peut entraver lanalyse e e
143
` g om trique pour aboutir a une appr hension op ratoire du dessin : cette e e e e ` derni` re consiste a savoir reconnatre des (sous-)congurations cl s, a travailler e e ` le dessin, le modier pour rendre plus visible une telle conguration. Des t ches de construction de gures g om triques, de reproduction de gures, a e e ` ou en sens inverse de description de gures en langage naturel contribuent a cet apprentissage. Pour construire ou d crire une gure, on doit s lectionner e e les propri t s qui la caract risent ; de telles activit s pr parent ainsi la ee e e e d monstration dans laquelle des propri t s donn es en impliquent dautres. e ee e LA GEOMETRIE DANS LESPACE
Si la g om trie sappuie sur des dessins plans, elle est cependant destin e e e e ` ` a mod liser lespace a trois dimensions. La lecture g om trique dun dessin e e e ` plan dun objet a trois dimensions pose des difcult s suppl mentaires aux e e e el` ves, puisque des evidences spatiales sur le dessin (comme la concourance de droites) ne traduisent pas n cessairement des propri t s g om triques e ee e e correspondantes. Il est unanimement recommand dans le monde entier de e ne pas n gliger la g om trie dans lespace et les probl` mes de repr sentation e e e e e dobjets de lespace, les r` gles de repr sentation prenant justement appui e e sur des propri t s g om triques. Des recherches sur lapprentissage de la ee e e g om trie 3D ont dabord port sur les capacit s de repr sentation dans le e e e e e plan dobjets solides, ou de lecture de dessins 2D dobjets 3D [Gaulin 1985 ; Bessot & Eberhard 1986 ; Parzysz 1991], puis sur des processus denseignement [Bessot & Eberhard 1987 ; Osta 1998 ; Douady & Parzysz 1998]. Il semble bien que les proc d s de repr sentation en perspective posent des probl` mes e e e e e durables aux el` ves, un peu partout dans le monde. Parzysz [1988] propose de les interpr ter en termes de compromis voir/savoir . En absence de e e ` connaissances g om triques, les el` ves cherchent a rendre compte dans le e e ` dessin a la fois de ce quils savent sur lobjet solide et de ce quils voient de ` ce dernier. De plus, leurs dessins peuvent chercher a reproduire des st r otypes ee de dessins familiers comme celui du cube en perspective cavali` re. Seul un e enseignement fond sur des connaissances g om triques permet de d passer e e e e ces difcult s. Seuls dix pour cent d l` ves de n d cole secondaire (1517 e ee e ans) afrment que lombre projet e sur un plan horizontal, par une lumi` re e e quasi-ponctuelle, dun carr fait de tiges rigides et parall` le a ce plan est e e ` e toujours un carr ; les autres el` ves h sitent lorsque la source de lumi` re nest e e e ` pas sur laxe vertical du carr . Cest le raisonnement a laide de th or` mes e e e sur les intersections de plans et droites, puis lintroduction de lhomoth tie, e qui permet de d passer les h sitations [Douady & Parzysz 1998]. e e
144
C. LABORDE
Or, dans les faits, trop souvent les chapitres de g om trie dans lespace sont e e les premiers laiss s de c t par les enseignants en cas de manque de temps. e oe On peut soupconner que les enseignants se sentent moins familiers avec cette g om trie souvent mal connue. Des questions consid r es comme triviales en e e ee g om trie plane ne recueillent pas de r ponses imm diates, telle la suivante : e e e e dans un t tra` dre, les hauteurs sont-elles concourantes ? . Le recours plus e e ` difcile a lexp rimentation en 3D explique cette absence de familiarit . Les e e nouvelles technologies facilitent lexp rimentation, comme le montrent les e gures 4 et 5 dun t tra` dre orthocentrique, modi ensuite par manipulation e e e directe dun de ses sommets dans un prototype de Cabri-g om` tre 3D. e e
FIGURE 4 Un t tra` dre orthocentrique et ses hauteurs e e
FIGURE 5 Apr` s modication dun des sommets e
La commission francaise de r exion sur lenseignement des math matiques e e sugg` re deux th` mes pour lenseignement de la g om trie dans lespace : les e e e e poly` dres et la g om trie sph rique. Comme on le voit, ces deux th` mes sont e e e e e dimportance pour une meilleure connaissance du monde dans lequel nous ` vivons et pour les applications a dautres disciplines qui utilisent les mod` les e fournis par la g om trie, ainsi que le souligne le rapport anglais mentionn au e e e d but de cet expos . Un autre apport de lenseignement de la g om trie dans e e e e lespace r side dans lusage de connaissances de g om trie plane auxquelles e e e toute activit de g om trie dans lespace fait appel. Cest justement le recours e e e e au jeu entre 2D et 3D qui fait le plus souvent cruellement d faut aux el` ves e de l cole secondaire. Des curricula comme ceux des Pays Bas, qui instaurent e une v ritable interaction entre plan et espace tout au long de la scolarit , sont e e de ce point de vue particuli` rement bienvenus. e
145
4. RAISONNEMENT
GEOMETRIQUE ET DEMONSTRATION
e Dans la droite ligne de la tradition des El ments dEuclide, la g om trie e e e a et consid r e, jusquil y a encore r cemment, comme le lieu dexercice ee e privil gi de la d monstration. On a pu sinterroger sur le bien-fond de e e e e e ` cette tradition pour initier les el` ves a la d duction. En effet, il sagit de e e faire comprendre aux el` ves quen g om trie, ils doivent saider de ce quils e e voient pour avoir des id es dans la r solution dun probl` me de g om trie, e e e e e mais quils nont plus le droit dutiliser ce quils voient quand il sagit de raisonner rigoureusement et doivent alors se cantonner au niveau th orique. e Un contrat didactique dun type nouveau doit etre install en classe et cela e nest pas simple. De nombreuses recherches au plan international ont montr combien e e lentr e dans la d monstration est difcile pour les el` ves de lenseignement e e secondaire voire universitaire, qui pr f` rent donner des arguments empiriques ee plut t que d velopper un raisonnement d ductif [Balacheff 1991 ; Harel o e e & Sowder 1998 ; Chazan 1993]. Ces arguments empiriques reposent sur des constats perceptifs de relations spatio-graphiques sur le dessin. La d monstration nest pas percue comme un outil de r solution de probl` me e e e ou dexplication et reste souvent une activit formelle d pourvue de sens e e [Hanna & Jahnke 1993 ; De Villiers 1994]. Les interpr tations de ces difcult s et les propositions denseignement ont e e pu soit envisager lactivit globale d laboration dune d monstration allant e e e ` de sa recherche a la production dun texte, soit se centrer particuli` rement sur e les sp cicit s du texte de d monstration. e e e Dans la seconde cat gorie de recherches, Duval [1991] a montr que e e le fonctionnement du texte de d monstration diff` re compl` tement du texte e e e argumentatif. Dans un texte de d monstration, ce qui importe nest pas la valeur e epist mique de v rit ou de vraisemblance dune proposition mais son statut e e e op ratoire (hypoth` se du probl` me, conclusion dun pas de d monstration), et e e e e ce qui permet le d roulement est la substitution dune ou de propositions par e ` une autre jusqu` ce quon aboutisse a la proposition cible. Une argumentation a en revanche repose sur laccumulation des arguments et leur force. Une des ` implications pour lenseignement est laccent a mettre sur lapprentissage de la r daction de la d monstration. La d monstration est un texte qui a un e e e fonctionnement discursif sp cique qui ne peut etre appris par simple imitation. e Alors que les travaux de Duval s parent la partie heuristique et la e r daction de la d monstration, dautres recherches ont au contraire vu le texte e e de d monstration comme indissociable de son elaboration et du probl` me e e
146
C. LABORDE
` e a r soudre. Ainsi des chercheurs italiens [Bartolini Bussi & Boero 1998 ; Bartolini Bussi et al. 1999 ; Douek 1999] voient-ils une continuit ou encore ce e quils appellent une unit cognitive entre la phase de production de conjectures e et celle de leur justication, le raisonnement qui conduit aux conjectures fournissant les arguments qui servent dans la construction de la preuve. Ces travaux mettent justement en evidence l volution des argumentations des e e el` ves, obtenues gr ce aux interventions de lenseignant et aux echanges entre a e el` ves organis s par lenseignant. Mais, comme ces chercheurs le soulignent, e le choix des probl` mes, et plus largement du th` me math matique sur lequel e e e porte l tude qui d bouche sur la construction de preuve, est fondamental. e e Ils consid` rent que le choix de faire produire des th or` mes (selon leur e e e e expression) par des el` ves est particuli` rement pertinent lorsque la g om trie e e e apparat comme un outil de mod lisation de ph nom` nes fondamentaux de e e e e lexp rience quotidienne des el` ves, telles les ombres produites par le soleil ou e la repr sentation en perspective. Cette approche conf` re une valeur culturelle e e ` a la g om trie qui d borde le strict plan math matique. Les conceptions e e e e e premi` res des el` ves elabor es dans le cadre de la vie quotidienne sont erron es e e e d` s quil sagit daller au-del` des premi` res evidences. La dimension th orique e a e e de la g om trie est ainsi construite comme permettant davoir un controle plus e e etendu sur ces ph nom` nes mal matris s. Deux caract ristiques de ces projets e e e e denseignement autour de la preuve en g om trie sont essentielles : e e
la gestion de la classe par lenseignant et ses interventions, en particulier dans les discussions collectives organis es dans la classe, qui contribuent e ` e a lavanc e dans la th orisation par les el` ves, e e le temps long des projets (plusieurs mois voire un an).
` La conception de situations denseignement contribuant a lapprentissage de la preuve est percue par tous, enseignants, formateurs denseignants et chercheurs, comme fondamentale. Les projets italiens fournissent une approche ` dans laquelle la preuve contribue a expliquer ou pr voir des ph nom` nes e e e externes aux math matiques. Dans une autre approche, la preuve est un moyen e ` de d passer lincertitude relative a une assertion math matique, comme dans les e e situations de validation de [Brousseau 1997]. Cette pluralit dapproches traduit e la diversit des fonctions de la preuve en math matiques et tr` s certainement e e e ` il importe que lenseignement re` te cette diversit et ne se restreigne pas a e e une vue etroite de la preuve dans laquelle il sagit de d montrer la validit e e d nonc s propos s par lenseignant. e e e De Villiers [1994] a soulign la diversit des fonctions dune preuve en e e g om trie et en particulier d fendu lid e que lenseignement pouvait prendre e e e e
147
davantage appui sur la fonction dexplication. La discussion sur ces fonctions est redevenue particuli` rement dactualit alors que les environnements de e e g om trie dynamique p n traient lenseignement. Certains ont pu en effet e e e e craindre lan antissement de la n cessit de lenseignement de la d monstration e e e e ` provoqu par ces environnements qui donnent a voir trop facilement nombre e de propositions tr` s probablement valides puisque v ri es dans la d formation e e e e e continue de la gure lors dun d placement quelconque dun de ses el ments. e Lexistence de ces nouvelles possibilit s de visualisation change profond ment e e ` e le rapport a l vidence sensible en g om trie et donc lenseignement de la e e g om trie. Le paragraphe qui suit est consacr a ces environnements. e e e ` 5. DES ` LOGICIELS DE GEOMETRIE DYNAMIQUE A MANIPULATION DIRECTE
Lid e de g om trie dynamique nest pas neuve. La propri t de la somme e e e ee des angles dun triangle est ainsi introduite par Clairaut par lid e que lors e du d placement de C sur le c t AC xe, la variation de langle B est e oe compens e par celle de C : e
FIGURE 6 Supposons, par exemple, que BC tournant autour du point B, s carte de AB, pour e sapprocher de BE , il est clair que pendant que BC tournerait, langle B souvrirait continuellement; & quau contraire langle C se resserrerait de plus en plus; ce qui dabord pourrait faire pr sumer que, dans ce cas, la diminution de langle C egalerait e laugmentation de langle B, & quainsi la somme des trois angles A, B, C , serait toujours la m me, quelle que f t linclinaison des lignes AC , BC , sur la ligne AE . e u [Clairaut 1741, premi` re partie LXII] e
148
C. LABORDE
Deux si` cles plus tard, a lieu une proposition faite en France par M ray e e ` [1874] denseigner la g om trie a partir du mouvement : le mouvement de e e translation permet dintroduire le parall lisme, le mouvement de rotation la e perpendicularit . Si M ray sest int ress au mouvement des gures solides e e e e ind formables [1906 (3e edition), p. 3], il evoque cependant [p. 7] la notion de e gure variable :
Une d formation lente et sans rupture, dun corps mou comme une p te e a ferme, ou exible comme un ressort, [] nous montre une succession de plusieurs gures solides in gales mais lorigine commune qui caract rise toutes e e ces gures, les ressemblances plus ou moins accentu es qui sont observables e entre elles, permettent de voir dans le ph nom` ne une m me gure, que le e e e d placement de ses divers points dans lespace fait varier dans sa forme. e [M ray 1906, 7] e
Depuis une quinzaine dann es sont apparus dans diff rents pays du monde e e des logiciels de g om trie dynamique 6 ) qui r alisent cette id e de gure e e e e variable, en offrant des entit s graphiques sur l cran de lordinateur ou de la e e calculatrice qui sont le r sultat dune suite dop rations exprim es en termes de e e e g om trie ; e e que lon peut d placer en manipulation directe avec la souris ; e et dont le comportement au cours du d placement est contr l par e oe une th orie g om trique (celle sous-jacente au logiciel) ; les relations e e e g om triques ayant servi dans le programme de construction du dessin e e sont conserv es au cours du d placement [Laborde 1995]. e e ` e En un mot, les dessins a l cran, une fois construits, ont un comportement qui ne suit pas forc ment les d sirs de leur auteur. Dans les modications e e e ` continues quentrane le d placement dun el ment du dessin, sont offertes a e la perception des propri t s spatiales dont on sait que l volution est r gul e ee e e e par la g om trie. On pourrait comparer ces r alit s spatio-graphiques dun e e e e nouveau type aux objets du monde r el, en disant quils r sistent aux manie e pulations de lindividu en suivant les lois de la g om trie. Lobservation de e e ` leurs comportements peut etre a lorigine de sp culations sur des relations e g om triques satisfaites par le r f rent g om trique correspondant. En partie e ee e e culier, des propri t s spatio-graphiques invariantes au cours du d placement ee e ` sont de tr` s bonnes candidates a etre des relations g om triques. La carace e e t ristique de tels logiciels est d tablir un lien entre les deux registres de la e e
6 ` ) Citons ceux qui ont donn lieu a des recherches internationales : Cabri-g om` tre (Laborde e e e & Straesser, 1990), Geometers Sketchpad (Key Curriculum Press, 1993), Geometry Inventor (Arcavi & Hadas, 2000)
149
g om trie. On peut cependant craindre que les apprenants d butants ne dise e e posant pas dune mobilit contr l e entre spatio-graphique et g om trique ne e oe e e sachent utiliser pleinement lamplication que permettent ces logiciels dans la visualisation de propri t s spatiales. ee Plusieurs recherches et projets denseignement au niveau de lenseignement secondaire ou n d cole primaire se sont donn comme objectif de mettre e e au point des situations dapprentissage de cette mobilit , en tirant parti e des possibilit s offertes par ces environnements. Ces derniers permettent de e concevoir de nouveaux types de t ches dans lesquelles les connaissances a ` g om triques servent a (re)produire, expliquer, pr dire des ph nom` nes spatioe e e e e graphiques constat s visuellement. e La (re)production de ph nom` nes spatio-graphiques avec le logiciel core e ` respond a des probl` mes de construction g om trique dans lesquels les e e e sp cications dun objet sont donn es soit discursivement, soit sous forme e e dun dessin dynamique. Lobjet reproduit doit satisfaire aux sp cications e et le d placement est un outil fort dinvalidation de proc d s de construce e e e tion au jug , fr quents chez les el` ves comme dit plus haut. Le contrat e e selon lequel ce nest pas le dessin produit qui int resse lenseignant mais e le proc d dobtention de ce dessin, ne semble plus tirer son origine dune e e demande de lenseignant mais de r` gles dusage du logiciel. Cela ne signie e e pas quil soit imm diatement appropri par les el` ves [Bellemain & Capponi e e 1991 ; Str sser 1992]. Diverses recherches internationales ont mis en evidence a e nombre de proc d s de construction chez les el` ves m langeant le jug et e e e e lusage de propri t s g om triques [Noss et al. 1994 ; Jones 1998 ; Holzl 1996 ; ee e e ` Healy 2000]. La distinction a etablir entre spatio-graphique et g om trique e e ` e est longue a construire par les el` ves pour quelle soit v ritablement approe pri e. Les environnements de g om trie dynamique apparaissent comme un e e e moyen de m dier la distinction spatio-graphique g om trique par la r sistance e e e e e au d placement et offrent des r troactions aux constructions des el` ves leur e e montrant linad quation de leurs constructions au jug . En cela ils contribuent e e ` a lapprentissage de cette distinction. Les t ches de pr diction sont particuli` rement adapt es aux logiciels de a e e e e g om trie dynamique. Il sagit pour les el` ves de pr dire le comportement e e e d l ments de la gure dans le d placement. Pour etre correct, cette pr diction ee e e prend appui sur des savoirs g om triques. Par exemple, comment se d place e e e le cercle image dun cercle donn dans une homoth tie lorsque le centre e e e dhomoth tie se rapproche du cercle donn ? (La plupart des el` ves donnent e e e une r ponse erron e.) La confrontation entre les pr dictions des el` ves et le e e e tre la source dun conit cognitif qui sav` re comportement observ peut e e e
150
C. LABORDE
moteur dans lapprentissage de la g om trie. e e Lexplication de ph nom` nes visuels peut etre sollicit e lorsque ces derniers e e e e e sont surprenants pour les el` ves. Donnons un exemple avec des el` ves francais de 1516 ans utilisant le logiciel Cabri-g om` tre [Clarou et al. 2001]. Deux e e vecteurs MA et MB dorigine commune M sont repr sent s a l cran. On e e ` e construit, avec loutil somme de vecteurs, la somme MC de MA et MB e et lon d place le point M . Si lenseignant demande aux el` ves, de dire ce e quils constatent de remarquable, ils ne savent que r pondre : la plupart du e temps, ils nont rien vu . Lexistence dun point invariant par lequel passent tous les vecteurs sommes dorigine M nest pas remarqu e, en absence de e mat rialisation de ce point. La situation devient tout autre d` s que lon d place e e e M en laissant la trace du vecteur somme (outil Trace). Le point invariant ` apparat comme un etranglement de la forme a deux dimensions cr ee par la e e trace (Fig. 7). Tous les el` ves le remarquent et leur surprise est grande. Cest e le moment ad quat pour lenseignant pour demander aux el` ves dexpliquer e ce ph nom` ne etonnant. Les connaissances th oriques servent alors doutil e e e dexplication dun ph nom` ne visuel. e e

FIGURE 7
Loutil Trace a permis la visualisation du point invariant car il la plac e au sein dune r gion de dimension 2, alors quauparavant il n tait que sur un e e objet de dimension 1, le vecteur somme. Le logiciel a servi damplicateur visuel de ph nom` nes math matiques. e e e La d monstration peut ainsi prendre sens dans de tels environnements si elle e remplit une nouvelle fonction, celle dexpliquer des ph nom` nes surprenants e e e pour les el` ves. Dans le m me esprit, des t ches de construction dans ces e a logiciels dobjets impossibles (par exemple, triangle avec deux bissectrices int rieures perpendiculaires, quadrilat` re avec trois angles droits) sont en e e
151
e g n ral des moments math matiques forts pour les el` ves, par le sentiment e e e de conviction que cr e lenvironnement de limpossibilit de ces objets, en e e ` e ` opposition a lattente des el` ves. Le recours a la d monstration peut ainsi etre e ` introduit par lenseignant dans la premi` re t che pour r pondre a la question e a e e du pourquoi spontan ment pos e par la majorit des el` ves. En effet, seul le e e e recours au th orique peut expliquer une non-existence puisquil est impossible e ` dexhiber un objet satisfaisant aux conditions. Il est a noter que ce peut etre loccasion dintroduire un raisonnement par labsurde. La seconde t che a e utilis e a l cole el mentaire favorise la prise de conscience du caract` re e ` e e de n cessit que poss` dent certaines propri t s g om triques d` s lors que e e e ee e e e dautres sont v ri es. Le caract` re apodictique des math matiques nest pas e e e e e percu par les el` ves d cole primaire ou de d but denseignement secondaire e e qui voient bien davantage les propri t s dun objet g om trique comme des ee e e qualit s ind pendantes (un carr a quatre c t s egaux, il a quatre angles droits, e e e oe etc.). Or la prise de conscience de ces liens de n cessit est un constituant e e indispensable de lapprentissage de la d monstration. Les environnements de e g om trie dynamique, par le sentiment de certitude quils offrent, peuvent etre e e utilis s a cette n. e ` Le probl` me pour lenseignant est didentier les situations g om triques e e e ` qui conduisent a ce sentiment de surprise. La conception de telles situations e qui n cessitent une grande connaissance des el` ves nest donc pas facile et e requiert du temps et de lexp rience. Il a fallu quelques si` cles et beaucoup e e de g om trie pour que le silicium du sable sur lequel etaient trac s les dessins e e e e e eph m` res dArchim` de concr tise les exp riences de pens e de Clairaut et de e e e e M ray. Le d pour lenseignement est quelles ne deviennent pas des r alit s e e e e e si` cle. e virtuelles banales pour les el` ves de la n du 21 e
REFERENCES ARGAUD, H.-C. Probl` mes et milieux a-didactiques pour un processus dapprentissage e ` e e en g om trie plane a l cole el mentaire, dans les environnements papier crayon e e et Cabri-g om` tre. Th` se de luniversit Joseph Fourier, Grenoble, 1998. e e e e ARSAC, G. V rit des axiomes et des th or` mes en g om trie v rication et e e e e e e e d monstration. Petit x 37 (1993), 533. e Axiomatique de la g om trie. Editions Al as & IREM de Lyon, Lyon, 1998. e e e BALACHEFF, N. Treatment of refutations : aspects of the complexity of a constructivist approach to mathematics learning. In : E. von Glasersfeld (ed.), Radical Constructivism in Mathematics Education, 89110. Kluwer, Dordrecht, 1991. BARTOLINI BUSSI, M. and P. BOERO. Teaching and learning geometry in contexts. In : [Mammana & Villani 1998], chap. 2, III, 5261.
152
C. LABORDE
BARTOLINI BUSSI, M. G., M. BONI, F. FERRI and R. GARUTI. Early approach to theoretical thinking : gears in primary school. Educational Studies in Mathematics 39 (1999), 6787. BELLEMAIN, F. et B. CAPPONI. Sp cicit de lorganisation dune s quence denseie e e gnement lors de lutilisation de lordinateur. Educational Studies in Mathematics 23 (1992), 5997. BERTHELOT, R. and M.-H. SALIN. The role of pupils spatial knowledge in the elementary teaching of geometry. In : [Mammana & Villani 1998], chap. 2, V, 7178. ` BESSOT, A. et M. EBERHARD. Adaptation de la perspective a une situation complexe e par des el` ves de 912 ans. European Journal of Psychology of Education 1 (1986), 8396. BESSOT, A. et M. EBERHARD. Repr sentations graphiques et th orisation de lespace e e des polycubes. Un processus didactique. In : Didactique et acquisition des connaissances scientiques. Actes du colloque de S` vres (Mai 1987), eds. e Vergnaud et al., 87108. Ed. La Pens e Sauvage, Grenoble, 1987. e BROUSSEAU, G. Etudes de question denseignement. Un exemple : la g om trie. e e S minaire de didactique des math matiques et de linformatique LSD2IMAG e e (1982-1983), 183226. Univ. Joseph Fourier, Grenoble, 1983. Theory of Didactical Situations in Mathematics. Didactique des Math matiques e 19701990. Kluwer, Dordrecht, 1997. CHAZAN, D. High school geometry students justication for their views of empirical evidence and mathematical proof. Educational Studies in Mathematics 24 (1993), 359387. CHEVALLARD, Y. Enseignement insens , enseignement raisonn et cr ativit sociale. e e e e Bulletin de lAPMEP 435 (2001), 526539. CLAROU, P., C. LABORDE et B. CAPPONI. G om trie avec Cabri Sc narios pour le e e e lyc e. CRDP, Grenoble, 2001. e e CLAIRAUT, A. El ments de g om trie. Lambert & Durand, Paris, 1741. e e CREM (COMMISSION DE REFLEXION SUR LENSEIGNEMENT DES MATHEMATIQUES). [2000] Rapport d tape sur la g om trie et son enseignement. In : Rapport au e e e ministre de l ducation nationale sur lenseignement des sciences math matiques, e e sous la dir. de J.-P. Kahane. O. Jacob/CNDP, Paris, 2002. DE VILLIERS, M. D. The role and function of hierachical classication of quadrilaterals. For the Learning of Mathematics 14 (1994), 1118. DOUADY, R. and B. PARZYSZ. Geometry in the classroom. In : [Mammana & Villani 1998], chap. 5, 159192. DOUEK, N. Argumentation and conceptualization in context : a case study on sun shadows in primary school. Educational Studies in Mathematics 39 (1999), 89 110. DUVAL, R. Structure du raisonnement d ductif et apprentissage de la d monstration. e e Educational Studies in Mathematics 22 (1991), 233261. Geometry from a cognitive point of view. In : [Mammana & Villani 1998], chap. 2, II, 3752. FISHBEIN, E. The theory of gural concepts. Educational Studies in Mathematics 24 (1993), 139162. FREUDENTHAL, H. Mathematics as an Educational Task. Reidel, Dordrecht, 1973.
153
GAULIN, C. The need for emphasizing various graphical representations of 3 -dimensional shapes and relations. In : Proceedings of the Ninth International Conference for the Psychology of Mathematics Education, ed. L. Streeand, Vol. II, 5371. State University, Utrecht, 1985. HANNA, G. and H. N. JAHNKE. Proof and application. Educational Studies in Mathematics 24 (1993), 421438. HAREL, G. and L. SOWDER. Students proof schemes : results from exploratory studies. In : A. H. Schoenfeld, J. Kaput and E. Dubinsky (eds.), Research in Collegiate Mathematics Education, III, 234283. Amer. Math. Soc., Providence, R. I., 1998. HEALY, L. Identifying and explaining geometrical relationship : Interactions with robust and soft Cabri constructions. 24th Conference of the International Group for the Psychology of Mathematics Education (PME 24), 1-103117. Hiroshima, 2000. HILBERT, D. und S. COHN-VOSSEN. Anschauliche Geometrie. Springer, Berlin, 1932. Trad. anglaise : Geometry and the Imagination. Chelsea, New York, 1952. HOLZL, R. How does dragging affect the learning of geometry. International Journal of Computers for Mathematical Learning 1 (1996), 169187. HOUDEMENT, C. et A. KUZNIAK. Un exemple de cadre conceptuel pour l tude de e lenseignement de la g om trie en formation des matres. Educational Studies in e e Mathematics 40 (1999), 282312. HOYLES, C., D. FOXMAN and D. A. KUCHEMANN. A Comparative Study of Geometry Curricula. QCA, London, 2001. JONES, K. Deductive and intuitive approaches to solving geometrical problems. In : [Mammana & Villani 1998], chap. 2, VI, 7883. LABORDE, C. et B. CAPPONI. Cabri-g om` tre constituant dun milieu pour lapprene e tissage de la notion de gure g om trique. Recherches en Didactique des e e Math matiques 14 (1994), 165210. e LABORDE, J. M. Des connaissances abstraites aux r alit s articielles, le concept e e de micromonde Cabri. In : Environnements Interactifs dApprentissage avec Ordinateur, eds. D. Guin, J. F. Nicaud et D. Py, 2841. Eyrolles, Paris, 1995. MAMMANA, C. and V. VILLANI (eds). Perspectives on the Teaching of Geometry for the 21st Century : An ICMI Study. Kluwer, Dordrecht, 1998. MARIOTTI, M. A. Images and concepts in geometrical reasoning. In : R. Sutherland and F. Mason, Exploiting Mental Imagery with Computers in Mathematics Education, 97116. NATO ASI Series, Springer-Verlag, BerlinHeidelberg, 1995. e MERAY, Ch. Nouveaux el ments de g om trie. Savy, Paris, 1874. Nouvelle edition revue e e et augment e : Jobard, Dijon, 1903. Troisi` me edition : 1906. e e NOSS, R., C. HOYLES, L. HEALY and R. HOLZL. Constructing meanings for constructing : an exploratory study for Cabri-G om` tre. In : Proceedings of the 18th International e e Conference for the Psychology of Mathematics Education, 360367. University of Lisbon, Lisbon, 1994. OSTA, I. CAD tools and the teaching of geometry. In : [Mammana & Villani 1998], chap. 4, 3, 128144. PARZYSZ, B. Knowing vs. seeing, problems of the plane representation of space geometry gures. Educational Studies in Mathematics 19 (1988), 7992. Representation of space and students conceptions at high school level. Educational Studies in Mathematics 22 (1991), 575593.
154
C. LABORDE
ROUCHE, N. Comment repenser lenseignement de la g om trie. Bulletin de lAPMEP e e 430 (2000), 600612. La g om trie et la nature des choses. Rep` resIREM 42 (2001), 2953. e e e STRASSER, R. Didaktische Perspektiven auf Werkzeug-Software im Geometrie-Unterricht der Sekund rstufe I. ZDM, 24 (5) (1992), 197201. a THIBAULT, M.-F. and R. LABARRE. Some hyperbolic geometry with Cabri-g om` tre. e e In : Intelligent Learning Environments : The Case of Geometry, ed. J.-M. Laborde (NATO ASI Series, Series F : Computer and Systems Sciences, Vol. 117), 218230. Springer-Verlag, Berlin, 1996. e ` TURC, A. Introduction el mentaire a la g om trie lobatschewskienne. K ndig, Gen` ve, e e u e 1914.
REACTION by Nicolas ROUCHE
Rudolf Bkouche has shown how, due to the evolution of the sciences and techniques, due also to the development of projective and non-Euclidean geometries, there emerged around 1900 an empiricist view of geometry, and how this view affected teaching. Now, if the relationship between geometry and experience appears to be so strong, it seems natural to investigate what geometry owes to our perceptions, which are the primary sources of our experience. This question is important as far as teaching is concerned. As evidence of this importance, we need only consider, with Colette Laborde, how students see so many disputable things in geometrical drawings. On this point of perceptions and in the scope of this note , let me mention only the contribution of Ernst Mach (see especially [Mach 1900]), already at the beginning of the twentieth century. This author pointed out the narrow scope of the human beings sense organs. In the case of sight, for instance, we do not see things clearly if they are either too large or too small, too close or too distant, or simply badly oriented with respect to us. Mach explains the particular conditions under which two congruent (isometric) gures are recognized as such : this happens when both gures are in a frontal plane, at a reasonable distance from the eyes, and providing that they are mutual images of each other, either by a reection whose axis belongs to the symmetry plane of the observer or by a horizontal translation. In all other cases, the congruence is recognized with difculty or not at all. Now, recognizing the congruence of two gures is no negligible operation. It is probably the rst step of the human being towards geometry. What do we have at our disposal to ascertain congruence, when our sight fails ? Mach again gives the answer. We can bring the gures in a position of correct perception, in the sense explained above. We can also superimpose one gure over the other, and Rudolf Bkouche reminded us that this simple
156
N. ROUCHE
operation is a starting point for Euclidean geometry. Mach calls these actions mechanical operations. But the objects cannot always be transported that way. And in case they cannot, our only resort lies in some operations of the intellect : we begin to elaborate and apply theorems.
FIGURE 1
FIGURE 2
By way of example, let us consider two triangles as in Fig. 1. Suppose we are told that it is a photographic picture of two triangles situated in space in an unknown way. Are they congruent ? We do not know ! But if, as in Fig. 2, we are given some indications as to their dimensions, we succeed in concluding that they are congruent. This we infer by an intellectual operation, whatever way we proceed : for example by noticing that the faces of a cube are congruent and that each triangle is half such a face, or by observing that our triangles have equal sides in pairs. This example sufces to show that plane geometry, before being the geometry of the plane, is rst of all the geometry of plane objects whatever their disposition in space. This throws some light on the issue of fusion between plane and solid geometry as commented by R. Bkouche. But let us come back to this basic fact : geometry has its roots in the visual, tactile and kinaesthetic perceptions. This is something we should remember when we want to revise the way it is taught. And this shows, in passing, that revising the teaching of geometry is not a matter for mathematicians only. Let us now consider the fties and sixties. Referring in general to the so-called new maths, Geoffrey Howson emphasizes a most relevant social fact, namely that the inspiration for this reform came from university mathematicians. These were concerned essentially with gifted students in senior high schools at a time, moreover, when these schools were attended by a minority of the population. G. Howson asks the frightening question :
GEOMETRY
157
Did they see school mathematics as simply the rst step in an assembly line that would eventually turn out a mathematics graduate ? This circumstance resulted, in several countries, in an axiomatic form of teaching, using an excess of technical terms and an unrealistically large and varied subject matter. It is true that certain countries have been preserved from such exaggerations. On the other hand, this curriculum planned for the elite inspired subsequently, but only subsequently, the teaching dispensed to everyone, including the pupils of kindergartens and primary schools. However, already before the fties, Piaget had been active in the eld of elementary school (see especially [Piaget & Inhelder 1947]). As G. Howson reminded us, his assertions were taken to justify the teaching of geometry in the direction from the poorest to the richest structures, in particular by putting topology rst (see also [Piaget 1966]). It is commonly admitted nowadays that the work of Piaget on the psychological genesis of geometry was biased by too narrow a conception of the essence of mathematics. As demonstrated by Hans Freudenthal, the result was that Piaget, in his investigations on geometry, proposed to the children only such situations Freudenthal would have said phenomena , lacking in variety, as suited his structuralist thesis. One might conclude that, to revise the teaching of geometry, one should rst revise Piaget (if one dare say that in Geneva), starting from another basis, and of course relying on all relevant contributions from psychologists and psychomotricians, but also with the collaboration of mathematicians. These are too rarely interested in kindergarten and primary teaching. But if we accept the idea that each stage of geometry learning is rooted in all the preceding ones, then the problem in no way reduces itself, as many mathematicians believe, to producing the best rational discourse, appropriate for 12 or 13 year-old children. And mathematicians could add much relevance to the studies of the most elementary learning activities, as observed by Freudenthal. They did not collaborate enough at the time of Piaget. Now, arriving at the third contribution in this panel and having no sufcient knowledge of the geometric software packages a subject of major importance , I would like to say a few words about the important analysis by Colette Laborde of all the misunderstandings originating in the gures in geometry courses. This analysis will certainly open the eyes of many teachers. On the other hand I found it interesting that this analysis relies on a clear distinction between perceived and conceived spaces, the visual and the geometric, between the drawing (usually called gure), which is perceptive and global, and the discourse, which is theoretical and analytic (linear). Colette
158
N. ROUCHE
Laborde asserts that the teaching of geometry relies heavily on what one usually calls gures, the common use of which in geometry is to help in the heuristic phase of problem solving. She adds :
e Il sagit de faire comprendre aux el` ves quen g om trie, ils doivent saider e e de ce quils voient pour avoir des id es dans la r solution dun probl` me de e e e g om trie, mais quils nont plus le droit dutiliser ce quils voient quand il e e sagit de raisonner rigoureusement et doivent alors se cantonner au niveau th orique. 1 ) [Laborde 2003, 145] e
One may wonder whether the role of geometric gures (drawings) is limited to the heuristic activity. Do there not exist particular gures which, because they represent a situation with few degrees of freedom, offer clearly to the sight, as well as to the mind, the variants, innite in number, of this situation ? Should the well known Indian proof of the Pythagoras theorem, as sketched in Fig. 3, be reformulated away from the gure ? Or is it not a quality of this proof that it demands no such formal expression ?
FIGURE 3
This type of observation is not proper to geometry. To mention just another example, an arithmetic property is sometimes given a proof which is expressed by considering particular numbers. Such a proof is said to be paradigmatic. This procedure works only when the particular situation refers unambiguously to all the other situations which have to be encompassed, but if this is the case, the procedure works and satises the most demanding mind.
1 ) The duty of teachers is to let the students understand that, although in geometry they have to rely on what they see to get ideas on how to solve a problem, they no longer have the right to utilise what they see as soon as the question is to reason rigorously; and then they must conne themselves to the theoretical level.
GEOMETRY
159
Vygotski once wrote : The analysis of reality with the help of concepts occurs much sooner than the analysis of the concepts themselves [1934 (1997)]. Should we not admit that the analysis of reality with the help of concepts is a fundamental activity to be exercised without end, in particular during all the school years, and that the analysis of concepts, which is much more difcult, comes along when the analysis of reality with the help of concepts encounters difculties ? I would like to conclude by emphasizing the inventory proposed by Geoffrey Howson of the conditions of success of a curriculum reform. There are conditions of various orders, and each of them is probably necessary but certainly insufcient, and often also beyond our competence. In this perspective, I recall a quotation of L. Henkin which could be read about ten years ago on the front page of an important Belgian report on mathematical teaching :
Changing the mathematical teaching system is a long run. Those who will engage in that and try to complete it in a decennium will not succeed and will be entirely disappointed. Changing the education system, I know for sure that mathematicians cannot do it, teachers cannot do it, politicians cannot do it, it demands a global cooperation effort. [Minist` re de lEducation 1990] e
REFERENCES LABORDE, C. [2003] G om trie p riode 2000 et apr` s. This volume, 133154. e e e e MACH, E. [1900] Lanalyse des sensations, le rapport du physique au psychique. (Trad. F. Eggers et J.-M. Monnoyer.) Chambon, Nmes, 1996. [Die Analyse der Empndungen und das Verhaltnis des Physischen zum Psychischen (2e ed.). Fischer, Jena, 1900.] ` MINISTERE DE LEDUCATION, DE LA RECHERCHE ET DE LA FORMATION. Perspectives sur lenseignement des math matiques dans la Communaut francaise de Belgique. e e Bruxelles, 1990. PIAGET, J. Linitiation aux math matiques, les math matiques modernes et la psycholoe e gie de lenfant. LEnseign. Math. (2) 12 (1966), 289292. PIAGET, J. et B. INHELDER. La repr sentation de lespace chez lenfant. PUF, Paris, e 1947. VYGOTSKI, L. Pens e et langage. Trad. F. S` ve (original russe, Moscou, 1934). La e e Dispute, Paris, 1997.
GENERAL DISCUSSION (reported by Marta MENGHINI)
The discussion about geometry following Nicolas Rouches reaction divided into two areas, namely : the historical motivations that determined the reform movements in Europe and the United States during the 20th century, particularly in the second half of the century, leading to the reduction or elimination of classical Euclidean geometry from the curriculum ; the didactic and cognitive links between the perception of geometric properties and their proof. These two areas are closely related, since the latter was a major motivation for changes in the school mathematics curriculum, particularly in the second half of the 20th century. The comments of the participants in the discussion, however, are summarised here separately. HISTORICAL AND CULTURAL PERSPECTIVES. Geoffrey Howson and Andr e Revuz both addressed the rst point, reecting the situations in Great Britain and France respectively. The rationale for eliminating classical Euclidean geometry was apparently the same in both countries. To begin with, there was a general feeling that things were not quite right with education in general, and the same impression prevailed also in other countries. Then there was an awareness of the difculties inherent in presenting proofs at school level, particularly the difculty of deciding where a proof begins and how much is to be assumed. This is sometimes referred to as the starting-point problem. Commonly, axioms were perceived by students as absolute truths rather than as reasonable starting points. Yet, the rules of procedure were often not stated clearly enough to allow playing the game of proving.
GEOMETRY
161
However, the reactions to the situation were different in the two countries. England and Wales chose a path which, by chance, was partly similar to that recommended by Dieudonn . The approach to the reform of school e mathematics was not axiomatic in nature and, in fact, has sometimes been considered to be in the style of physics rather than mathematics. In France, initially, the problems of teaching mathematics were answered institutionally by setting up the IREM 1 ) groups, charged with the responsibility of both in-service teacher training and research into mathematics teaching practice. Only later did there arise a demand for major changes to the curriculum, which then adopted a Bourbakist approach, following the changes that were already under way in Belgium. The answer to the problem of axioms was to abandon the Euclidean postulates entirely and replace them by new axioms together with a different way of introducing them. During the same period in Italy there were no ofcial changes in the mathematics curriculum at all. In fact, the debate about geometry teaching had been initiated at the beginning of the century. At that time it focused on the interaction between empiricism and axiomatics. Aldo Brigaglia commented that this did not represent a contraposition of views. Peano, in fact, had been a supporter of empiricism, although he was rigidly axiomatic in his expositions. On the other hand, Cremona was not particularly keen about axiomatics in research, but his geometry was abstract, not empirical. Finally, Enriques axioms for teaching had a psychological genesis. During the 60s and 70s certain aspects of the European mathematics teaching reforms were adopted in Italy, by means of integrating Choquets axiomatics (which represented a way of introducing measurement and geometrical transformations) into the classical structure of geometry. The Italian cultural tradition did not allow for a disappearence of congruence and similarity criteria for triangles, as had occurred during a period of thirty years in France, where they were replaced for ten years by linear algebra and then by the more empirical geometric transformations. DIDACTIC AND COGNITIVE ASPECTS OF GEOMETRIC PROPERTIES AND THEIR Several speakers pointed out that today the attitude about the teaching of geometry has changed, particularly regarding the relationship between intuition and proof. Euclidean geometry without tools is not the same thing as Euclidean geometry with tools, and undoubtedly DGS (Dynamic Geometry
PROOF.
1
) Institut de recherche sur lenseignement des math matiques e
162
GENERAL DISCUSSION
Software) has a role in the revival of this debate. The topic generated a lively discussion, often in the form of open questions, such as : how is it possible to obtain a proof after visualizing a property with DGS ? is there a dichotomy between the gure and the formal explanation ? are axioms the bridges between visualization and proof ? But there were also questions of a more fundamental nature, such as : why, more generally, is it necessary to provide a proof for a property that looks evident ? what leads us to abstraction ? what is perceptive understanding with respect to proof ? Underlying the discussion was the recognition of the deductive nature of proof and its presence in a number of disciplines. Perception without (deductive) proof is unsatisfactory and some speakers held that geometry was the logical analysis of our perception of space. The question remains, however, as to what draws us to this analysis ? What is fundamental to the problem is not the existence of the two aspects of understanding intuition and proof , but the link between them. Without such a link, there would be two different disciplines : on the one hand the gure, reasoning on the other ; but then mathematics which is based entirely on reasoning would seem to shun empiricism. Colette Laborde expressed the opinion that a dichotomy between these two aspects should not be made, and that there is, on the contrary, a dialectic. Geometry cannot be done without both aspects. But for the purpose of analysing the problems that students encounter when doing geometry, it is interesting, and fruitful, for the observer to make a distinction between diagrams and theory. On the other hand, the nature of the dialectic between reasoning and empiricism needs to be explored. Gila Hanna, for example, pointed out that it is not always the case that in visualization we see that something is true, and in abstraction we explain why it is true. Abstraction can also be used to show that something is true, without ever showing why it is true : there are proofs that do not mention anything about explanations. On the other hand, the example of a proof without words given by Nicolas Rouche cannot be brought as an example of uselessness of a formal explanation. Such diagrams, inherently implying a proof, do not need a formal proof for a mathematician, because mathematicians are experts and are able to extract the relevant properties from the drawing. However, it cannot be assumed that 1213 year-old children would be able to understand. Also other types of proof, such as proofs by contradiction, question the role of visualization, since
GEOMETRY
163
it is necessary to use theory in these cases. As against that, visualization is certainly a powerful tool. In using DGS, for example, students are convinced, in less time than with paper and pencil, of the non-existence of an object (for example of a triangle with two perpendicular bisectors). Many speakers were of the opinion that new developments in dynamic geometry would lead more and more towards aiding the proof of geometrical results. Nonetheless, we still need to bear in mind that tools are only tools, and that other questions about mathematics may arise (such as the interplay between the continuous and the discrete). In this sense, DGS has more mathematics embedded in it than ruler and compasses alone. Finally, tools should be evaluated not for their perfection, but mainly for their appropriateness for teaching.
ANALYSIS
LENSEIGNEMENT DU CALCUL DIFFERENTIEL ET INTEGRAL ` ` AU DEBUT DU VINGTIEME SIECLE The teaching of calculus at the beginning of the twentieth century by Jean-Pierre KAHANE
An international conference of ICMI was held in Paris in April 1914. A detailed account was published in LEnseignement Math matique. One item in the conference e agenda was the inquiry that ICMI had conducted all over Europe on the results obtained from introducing calculus in the upper grades of secondary institutions. In the rst section, we discuss the rather optimistic report written by E. Beke [1914], which talks about mathematics as an instrument of progress. In spite of its eloquent conclusion on the usefulness of calculus in the perspective of promoting the study of scientic humanities, the situation in the various countries was quite diverse, and the divisions were even amplied in the aftermath of the First World War. The second section is devoted to the context of calculus teaching, i.e. exercices, text-books, and teacher training. We look more specically at the situation in France and in Hungary. In two appendices we reproduce two specimens of examinations taken by students at the Sorbonne, resp. in the upper grade of the lyc e, around 1900. e
LENSEIGNEMENT DU CALCUL DIFFERENTIEL ET INTEGRAL ` ` AU DEBUT DU VINGTIEME SIECLE par Jean-Pierre KAHANE
Cette communication est divis e en deux parties. La premi` re expose la e e situation et les th` ses en pr sence dapr` s le compte rendu de la conf rence e e e e internationale davril 1914. La source principale est le rapport dE. Beke [1914]. J voquerai aussi les contributions dE. Borel et de Ch. Bioche, et la e discussion. La seconde partie est consacr e au contexte (exercices, livres, e formation des enseignants), en insistant sur la France et la Hongrie. En compl ment, on trouvera des exemples de ce que lon pouvait attendre des e e ` e el` ves a l poque.
1. LA
CONFERENCE INTERNATIONALE DAVRIL
1914
Cette conf rence internationale de lenseignement math matique, soigneusee e ` ment pr par e, se tint a Paris du 1er au 4 avril 1914. Un compte rendu d taill e e e e en est donn dans LEnseignement Math matique [Beke 1914]. La conf rence e e e rassemblait plus de 160 participants venant de 17 pays ; les pays les plus repr sent s sont la France (82), la Hongrie (15), lAllemagne (14), la Suisse e e (12), la Russie (10). Lordre du jour comprenait deux points : A) les r sultats obtenus dans lintroduction du calcul diff rentiel et int gral e e e dans les classes sup rieures de lenseignement moyen, et e B) la place et le r le des math matiques dans lenseignement technique o e sup rieur. e Cest du point A) quil sera question ici. Le discours inaugural de G. Castelnuovo, parlant au nom de Felix Klein, pr sident de la CIEM, emp ch , indique les raisons du choix de Paris comme e e e si` ge de la conf rence : e e
LENSEIGNEMENT DU CALCUL DIFFERENTIEL ET INTEGRAL
169
De la France est partie en 1902 linitiative dune r forme organique de e lenseignement traditionnel des ecoles moyennes. Le plan d tudes de cette e epoque a introduit dune mani` re syst matique, avant les autres pays, les notions e e de d riv es et de fonctions primitives dans les programmes des lyc es. Il est e e e ` donc naturel de constater ici ce que lexp rience de dix ans a pu enseigner a e ce sujet. [Castelnuovo 1914, 190]
Emile Borel, lun des artisans de cette r forme de 1902, fait la conf rence e e inaugurale sur le th` me de ladaptation de lenseignement secondaire aux e progr` s de la science [Borel 1914]. Il constate dabord que, pour toutes sortes e de raisons, lenseignement secondaire ne peut evoluer que tr` s lentement , e et il insiste :
Toute modication trop brusque ou trop consid rable risque d tre f cheuse e e a pendant un temps assez long ; on peut m me afrmer dune mani` re presque e e absolue que toute modication est tout dabord nuisible et, pendant la p riode e dadaptation, entrane plus dinconv nients que davantages. [Borel 1914, 201] e
Cependant lenseignement secondaire nest pas immuable, il evolue, dans ` toutes les mati` res, et les math matiques ne sauraient etre a part : e e
[] les changements doivent etre lents ; mais peut- tre nest-il pas excessif e de penser quil est aussi absurde pour le professeur de math matiques de e lenseignement secondaire de paratre ignorer Galil e, Descartes, Newton et e Leibniz quil le serait pour le professeur de chimie dignorer Lavoisier, ou pour le professeur dhistoire de n gliger la R volution francaise. [Borel 1914, e e 207]
Au surplus, la g om trie analytique et le calcul diff rentiel et int gral e e e e e sont, au moins dans leurs el ments, plus proches de lexp rience commune e e que beaucoup de math matiques dites el mentaires. Les lecteurs de journaux, e lorsquils regardent un graphique,
font de la g om trie analytique sans le savoir ; parfois m me, en discutant sur e e e la rapidit plus ou moins grande des oscillations de ces graphiques et sur les e cons quences quon peut en tirer, ils font, sans le savoir, du calcul diff rentiel e e et du calcul int gral. [Borel 1914, 206] e
Enn, dit Emile Borel, la valeur educative des mati` res nouvelles ne le e c` de en rien aux mati` res anciennes, m me si la masse des professeurs ne e e e ` peut arriver du premier coup a une technique p dagogique aussi bonne pour les e mati` res nouvelles que la technique traditionnelle l tait pour les anciennes . e e En r sum , tout changement est mauvais pendant quon le r alise et [] e e e un changement, sil nest pas absurde, devient bon une fois quil est r alis e e depuis un certain temps [Borel 1914, 210]. ` Cest le rapport dE. Beke, professeur a luniversit de Budapest, qui e e e e constitue la charpente du point A) de lordre du jour. Il avait et pr par par
170
J.-P. KAHANE
un questionnaire et par des rapports partiels venant dAllemagne, dAustralie, dAutriche, du Br sil, du Danemark, des Etats-Unis, de France, de Hollande, e de Hongrie, des Iles Britanniques, dItalie, de Norv` ge, de Russie, de Serbie e et de Suisse. LEnseignement Math matique en publie un r sum [Beke 1914, e e e 222225], le texte complet [Beke, 245282] et le questionnaire qui la pr par e e [Beke, 283284]. Jen d gagerai seulement quelques-uns des points saillants. e Lintroduction du rapport est vibrante doptimisme. Les math matiques sont e la langue naturelle de la pens e et linstrument du progr` s ; on attend e e des sciences plus deffet que par le pass , pour la formation des esprits . Et e surtout ceci, quon ne peut lire sans emotion :
Je crois pouvoir afrmer, sans crainte de me tromper, que cest un haut id al e dinternationalisme qui nous a r unis ici. Nous avons senti que l ducation de e e la jeunesse na pas seulement pour but de former, daccrotre et de maintenir les forces vives dune nation et lesprit national, de doter du patrimoine commun les ouvriers actifs de la civilisation nationale ; elle a aussi la t che encore a ` plus noble de cr er et de faire vivre un id al commun a toute lHumanit . e e e [Beke 1914, 246]
Nous sommes le 2 avril 1914. Dans quelques mois, la guerre. La conclusion du rapport est egalement eloquente, sur la n cessit de e e promouvoir des humanit s scientiques et sur la place dans ce cadre de e lobjet du d bat, le calcul diff rentiel et int gral : il faut e e e
faire r pandre dans le cercle le plus large possible, parmi tous les hommes qui e cultivent la Science, la connaissance du Calcul innit simal qui est la Science e du changement, principe eternel du monde, qui est linstrument indispensable de tout raisonnement scientique et qui, enn, repr sente une cr ation magnique e e de lesprit humain. [Beke 1914, 282]
Allons directement au questionnaire. Il concerne les lyc es, les gymnases e classiques ou r aux, les etablissements similaires des divers pays et aussi, en e principe, les ecoles normales dinstituteurs (mais il ny a pas de r ponse de e ce c t ). oe 1) Qua t-on introduit ? Le calcul diff rentiel est-il limit aux fonctions e e dune variable ? Quelles fonctions etudie-t-on ? Quid du calcul int gral ? de la e formule de Taylor ? des equations diff rentielles ? e 2) Quel est le degr de rigueur ? Se contente-t-on dune introduction e g om trique ? Utilise-t-on la notion de limite ? D montre-t-on des th or` mes e e e e e sur les limites, tels que lim(1 a) 1 lim a ? Fait-on usage des diff rentielles, e et comment ? Fait-on intervenir le reste dans la formule de Taylor ? Parlet-on de fonctions non-d rivables ? Introduit-on rigoureusement les nombres e irrationnels ?

171
3) Quelle est la m thode ? A-t-on d j` vu des fonctions simples et leur e ea repr sentation graphique ? Emploie-t-on la notation de Leibniz ? Commencee t-on par le calcul diff rentiel ou par le calcul int gral ? Comment introduite e on lint grale : int grale d nie (limite dune somme) ou int grale ind nie e e e e e (primitive) ? Quels manuels, quels ouvrages ? 4) Quelles applications ? Par exemple, quid des maxima et minima ? des d veloppements en s ries enti` res ? de linterpolation, de lextrapolation, du e e e calcul des erreurs ? du calcul des aires et des volumes ? de la m canique e (vitesse, acc l ration, travail, moments dinertie, etc.) ? de la physique, en ee particulier pour loptique et l lectrodynamique ? e 5) Quels all` gements par ailleurs ? e 6) Quel accueil pour la r forme ? e Le rapport de Beke donne une analyse d taill e des r ponses. Ce qui frappe e e e est la diversit des m thodes, en particulier dans lintroduction de lint grale. e e e En g n ral cependant, on se borne aux fonctions dune variable, on pr f` re e e ee ` la notation de Lagrange a celle de Leibniz, on ne parle pas de fonctions nond rivables, et les nombres irrationnels ne sont introduits que sur des exemples e (extraction de racines). La s rie de Taylor gure dans peu de programmes. e Les applications principales sont la recherche des extrema et les calculs daire et de volume. Mais ce r sum de r sum doit etre compl t , imm diatement, par un e e e e ee e apercu sur la diversit des situations. En Allemagne, les r ponses sont e e diff rentes selon quelles proviennent de Bavi` re, Wurtemberg, Bade et e e e Hambourg, o` les el ments du Calcul innit simal sont au programme des u e ` ecoles, ou de Prusse et de Saxe ou ils se trouvent enseign s sans etre au e programme. Un cas extr me est celui de Hambourg. Les d riv es y sont e e e ` enseign es depuis 1874, et aussi, a tire optionnel, les int grales. En 1897, les e e int grales font partie du programme egalement. En Hongrie, depuis 1899 la e plupart des lyc es enseignent les d riv es et les int grales ; depuis longtemps e e e e etaient dans les programmes la g om trie analytique, les repr sentations e e e graphiques de fonctions simples, les maxima et minima. Les Anglais insistent sur la qualit de leurs livres. En Italie, rien ; pas danalyse dans les lyc es e e (cependant quavec Dini, Peano, Volterra et dautres, lItalie joue un r ole phare en analyse au plan de la recherche). En Roumanie, plusieurs variables. En Suisse, une particularit m rite attention. Voici une r solution de lAssociation e e e suisse des professeurs de math matiques, en 1904 : e
La notion de fonction et les probl` mes fondamentaux qui sy rattachent e appartiennent au programme de lenseignement math matique des ecoles e moyennes. [Beke 1914, 256]
172
J.-P. KAHANE
Cest moi qui ai signal appartiennent. Ils ne disent pas devraient ape ` partenir, en sadressant a une instance sup rieure. Ils d cident 1 ). e e Je nai pas encore parl de la situation en France. Beke en parle e abondamment, et cest lobjet dun rapport sp cial de Charles Bioche [1914], e professeur au lyc e Louis le Grand. e Ce qui est le plus sp cial en France, cest lexistence des classes de e Math matiques sp ciales. Le calcul int gral est renvoy a ce niveau. Au lyc e, e e e e` e on se borne donc, avant le baccalaur at, au calcul diff rentiel : cest plus e e quen Italie, mais beaucoup moins qu` Hambourg ou Budapest. La r forme a e de 1902 a introduit les d riv es en classe de seconde (1415 ans), mais la e e ` r cente r forme de 1912 reporte cette introduction a la classe de premi` re e e e (1516 ans). Les d riv es des fonctions trigonom triques sin x , cos x , etc. e e e sont renvoy es en classe de Math matiques (1617 ans). e e Bioche donne des exemples d preuves de math matiques propos es dans e e e ` les classes de premi` re et de Math matiques. Jy reviendrai a la n de cet e e article. La discussion apporte des indications compl mentaires o` eclatent les e u diff rences entre les diff rents pays. Dans lensemble, les participants ape e prouvent la r forme. Une forte voix discordante est celle du math maticien e e autrichien R. Suppantschitsch, qui ecrit 2 ) : Die Frage, ob die Einf hrung der u Innitesimalrechnung einstimmig als ein entscheidender Fortschritt zu betrachten sei, ist mit NEIN zu beantworten. (Est-ce un progr` s ? NON !) e Et il ajoute : Ich halte daher schon jetzt die Frage fur sehr diskutierbar, ob die Innitesimalrechnung AUS der Schule nicht wieder verschwinden soll. (Il est temps de discuter de renvoyer le calcul innit simal DEHORS.) e Lopinion de Suppantschitsch apparaissait alors comme extr mement e minoritaire. Cest elle, cependant, qui triompha en France apr` s la guerre e de 19141918. e La fracture de la guerre a et non seulement la ruine des belles esp rances exprim es par Beke, mais un s v` re coup darr t a la r exion sur e e e e e ` e lenseignement des math matiques. La CIEM d chir e, divis e, ne sen est e e e e jamais remise.
1 2
) On trouvait une situation analogue en Allemagne, comme d crit dans [Schubring 2003, 53]. e ) LEnseign. Math. 16 (1914), 305.
173
2. CONTEXTE,
COMMENTAIRES ET COMPLEMENTS
Le d but du vingti` me si` cle est une p riode faste pour la recherche e e e e math matique. De jeunes math maticiens apparaissent sur le devant de la e e ` ` sc` ne : Borel, Baire, Lebesgue, Fej r, Riesz, pour men tenir a lanalyse et a e e la France ou la Hongrie. Lenseignement sup rieur suit tr` s in galement. Cest en France quil e e e ` manifeste le plus de retard. Les epreuves de calcul diff rentiel et int gral a la e e Sorbonne, comme celles danalyse sup rieure, sont int ressantes et difciles, e e mais ne se distinguent gu` re, jusque vers 1950, de celles qui etaient propos es e e ` a la n du 19e si` cle. Par exemple, lint grale de Lebesgue en est absente. e e Cest en Russie, en Pologne, en Hongrie, que lanalyse telle quelle est issue des travaux de Borel, Baire et Lebesgue sera enseign e et popularis e. e e ` Cependant l dition math matique est orissante, a tous les niveaux. Le e e Jahrbuch uber die Fortschritte der Mathematik en donne une image au plan mondial. En France existent dexcellents ouvrages denseignement, et, parmi ` e eux, des livres dexercices a lintention des el` ves et des professeurs. ` Lanalyse y apparat souvent a lombre de lalg` bre. Carlo Bourlet a publi e e e en 1896 ses Lecons dalg` bre el mentaire [Bourlet 1896]. Il y introduit les e ` e d riv es et les applique a l tude de la variation des fonctions, avant de passer e e aux logarithmes et aux int r ts compos s (facon de concevoir lexponentielle), ee e puis, en appendice, dintroduire les nombres complexes et l tude des fonctions e circulaires. Plus de 250 exercices accompagnent ces Lecons. Bien plus tard, en 1915, un Pr cis dalg` bre pour les classes de premi` re e e e C et D, de J. Girod [1915], comprend les formules g n rales pour le calcul e e des d riv es et une foule dapplications. Par exemple [chap. 6, th or` me 3] : e e e e si x 0 , y 0 ont une somme constante, le produit x p y q est maximum ` quand x et y sont proportionnels a leurs exposants . Les 200 exercices que contient louvrage sont en majorit des sujets dexamen donn s au e e baccalaur at : beaucoup de questions danalyse pr sent es comme probl` mes e e e e de g om trie. e e Lanalyse apparat sous son nom dans un livre de Baire de 1907, Lecons sur les th ories g n rales de lanalyse [Baire 1907]. Cest un excellent ouvrage, e e e tr` s rigoureux, mais qui sarr te aux fondements. e e Lanalyse classique, celle qui senseigne dans les universit s, fait lobjet e dun recueil, par E. Fabry, en 1913, de probl` mes danalyse math matique e e [Fabry 1913]. Le livre comprend 279 enonc s et solutions. Les rubriques e sont : quadratures, int grales multiples, fonctions analytiques, equations diff e e quations rentielles, courbes planes, surfaces et leurs lignes remarquables, e

174
J.-P. KAHANE
diff rentielles et leurs applications g om triques, diff rentielles totales. Cela e e e e donne une id e de la culture math matique des professeurs de lyc e les plus e e e solidement form s. e Il est int ressant de comparer la contribution des math maticiens les e e ` plus eminents a la formation des professeurs de lyc e dans trois pays bien e repr sent s a la conf rence de Paris : la France, lAllemagne et la Hongrie. e e ` e En France, lexemple est Jacques Hadamard, avec ses excellentes Lecons e de g om trie el mentaire [Hadamard 1898/1901]. Il sagit bien de lecons e e e adress es aux professeurs, mais en principe elles sadressent aussi aux el` ves. e Le jeu est bien de faire un expos el mentaire de la g om trie. Cela sera le parti ee e e pris egalement de Lebesgue, bien plus tard, dans ses cours sur les coniques. Ouvrages s duisants, ing nieux, et d lib r ment eloign s du mouvement des e e e ee e math matiques contemporaines. e ` En Allemagne, Felix Klein sadresse aux professeurs a partir dune e conception tr` s diff rente. Il sagit daborder les math matiques el mentaires e e e dun point de vue moderne : Elementarmathematik vom hoheren Standpunkte aus [Klein 1908/1909]. Par exemple, pourquoi y a-t-il des nombres transcendants ? Parce que lensemble des nombres alg briques est d nombrable, e e et que lensemble des nombres r els ne lest pas ; cest la d monstration de e e ` Cantor. Felix Klein sattache a eclairer le champ du point de vue le plus e elev possible. Son inuence internationale est consid rable il pr side e e e la CIEM, il a et linitiateur de lUnion math matique internationale. Mais e surtout, son audience en Allemagne est immense. Son jubil a G ttingen, en e ` o d cembre 1918, sera la premi` re manifestation scientique en Allemagne au e e sortir de la guerre. ` Revenons a Hadamard. Il est plus jeune que Klein, mais cest un math maticien de la m me envergure. Lui aussi sera, plus tard, pr sident e e e de la CIEM. Cest un passionn denseignement. Mais, pour lui, il y a rupture e compl` te entre sa recherche et lenseignement. Klein, lorsquil sadresse aux e e enseignants, sinspire le cas ech ant de son programme dErlangen. Hadamard, qui pr side la s ance o` parle Beke, est le sp cialiste mondial de la s rie de e e u e e ` Taylor et de son prolongement analytique. Y a-t-il quelque chose a en tirer pour lenseignement ? Non, dit-il, la s rie de Taylor na pas sa place au e lyc e , et cela cl t le d bat. e o e En Hongrie, le lien entre recherche et enseignement a pris des formes ` originales. Il existe une comp tition a la fois tr` s s lective et tr` s populaire e e e e ` dans le milieu math matique, ouverte aux etudiants qui entrent a lUniversit . e e Cest la comp tition Schweitzer, qui existe toujours, et dont les organisateurs e
175
comme les laur ats sont parmi les meilleurs math maticiens du pays. Il e e existe une revue destin e aux professeurs de lenseignement secondaire et aux e etudiants, r dig e par les math maticiens les plus cr atifs. Cest Matematikai e e e e Lapok. Leopold Fej r (Fej r Lipot), qui a pass sa th` se en 1902, introduit e e e e ` une pratique a laquelle il sera d` le sa vie durant : tout ce quil publie en e francais ou en allemand est egalement publi en hongrois. Tout cela cr e une e e relation tr` s forte entre chercheurs, professeurs et etudiants, et le hongrois e est une langue en etat de marche au plan scientique autant que les grandes langues de communication. Le poids qua E. Beke dans la conf rence, ce e nest pas seulement davoir dress un etat des lieux clair et complet ; cest e ` aussi de repr senter a la conf rence une tradition remarquable de liaison entre e e recherche, enseignement, et culture nationale. Pour conclure, il ne me semble pas mauvais de montrer ce qui etait e demand aux etudiants duniversit et aux el` ves de lyc e en France au e e e cours des ann es 1900. On trouvera donc ci-apr` s les sujets dexamen en e e ` calcul diff rentiel et calcul int gral a Paris en 1901. Comme je lai d j` e e ea dit, le type de sujets na pas beaucoup vari pendant cinquante ans. Mais il e nest pas evident que les etudiants de licence daujourdhui sachent les traiter (annexe 1). Lannexe 2 est emprunt e a lexpos de Ch. Bioche. Cest le e ` e ` sujet dune composition de math matiques, a traiter en 2 heures et demie, e e pour des el` ves de la classe de Math matiques (1617 ans). La forme de e e l nonc montre que les el` ves sont habitu s a repr senter certaines gures de e e e ` e r volution par leur section m ridienne. Peu de connaissances sont n cessaires ; e e e les candidats bacheliers daujourdhui sauraient-ils les mobiliser ? Mais nest-il ` pas vrai qu` chaque epoque les performances des an s apparaissent a la fois a e comme un peu etranges et ind passables ? e
176
J.-P. KAHANE
ANNEXE 1 Examens de la Sorbonne, juillet 1901 Certicat de Calcul diff rentiel et Calcul int gral e e
` Epreuve ecrite (2 juillet, de 8h. a midi) 1o Une famille de courbes gauches (T) est repr sent e, dans un syst` me e e e daxes rectangulaires, par les equations x2
et repr sentent deux constantes arbitraires. On demande 1) de e o` u d montrer que ces courbes sont les trajectoires orthogonales dune famille e de surfaces (S), d pendant dun param` tre arbitraire, et de trouver l quation e e e g n rale de ces surfaces ; 2) de prouver que les sections des surfaces (S) par e e des plans passant par Oz sont des lignes de courbure de ces surfaces et de trouver la seconde famille de lignes de courbure. ` 2o Soit R(x y) un polyn me entier a deux variables x y , de degr p q 1 o e ` au plus par rapport a lensemble des deux variables. D montrer que lint grale e e double p ((x2 a2 )p ) q ((y2 b2 )q ) I R(x y) dx dy xp xq ` etendue a lint rieur du rectanle limit par les droites x e e a, x a, y b, y b est nulle, quels que soient les coefcients du polyn me R(x y) . o
1 1 ` Epreuve pratique (2 juillet, de 2h 2 a 3h 2 ) Calculer lint grale d nie e e

x2 (1 x) dx (1 x)3 0 e soit par des moyens el mentaires, soit en se servant de la th orie des r sidus. e e Examinateurs : MM. Goursat, Picard, Hadamard. e Sur 37 candidats, 13 ont subi les epreuves orales et 11 ont et recus.
y2
z2
x2
2y2
z2 z
177
ANNEXE 2 Composition de math matiques e (classe de Math matiques, 2h.30) e
On consid` re le solide form par un c ne SAA et un cylindre ABB A e e o ayant la m me longueur de g n ratrice : SA AB a . Soit x la hauteur SH e e e du solide. 1o Exprimer le volume V du solide au moyen de a et de x . 2o Trouver pour quelles valeurs de x le volume V est maximum. Calculer ` ce maximum en hectolitres dans le cas ou a 1 m . 3o Construire la courbe qui repr sente les variations de la fonction e 3V a3 en repr sentant par a lunit de longueur graphique. e e o Calculer laire comprise entre la courbe et la corde joignant le point 4 dabscisse 1 et le point dabscisse 2 . 5o D duire de la consid ration de la courbe combien il y a de valeurs de e e x pour lesquelles y prend une valeur donn e. Calculer les valeurs de x qui e ` correspondent a y 3 .

[Bioche 1914, 287288]
178
J.-P. KAHANE
BIBLIOGRAPHIE BAIRE, R. Lecons sur les th ories g n rales de lanalyse. Gauthier-Villars, Paris, 1907. e e e BEKE, E. Les r sultats obtenus dans lintroduction du calcul diff rentiel et int gral e e e dans les classes sup rieures des etablissements secondaires. Rapport g n ral. e e e LEnseign. Math. 16 (1914), 222225 ; 245284. BIOCHE, Ch. Lorganisation de lenseignement du calcul des d riv es et des fonctions e e primitives dans les lyc es de France et sur les r sultats obtenus. LEnseign. Math. e e 16 (1914), 285289. BOREL, E. Ladaptation de lenseignement secondaire aux progr` s de la science. e LEnseign. Math. 16 (1914), 198210. e BOURLET, C. Lecons dalg` bre el mentaire. Colin, Paris, 1896. e ` CASTELNUOVO, G. Discours inaugural a la s ance douverture de la Conf rence e e internationale de lenseignement math matique (Paris, 14 avril 1914). LEnseign. e Math. 16 (1914), 188191. FABRY, E. Probl` mes danalyse math matique. Hermann, Paris, 1913. e e GIROD, J. Pr cis dalg` bre pour les classes de premi` re C et D. Alcan, Paris, 1915. e e e e HADAMARD, J. Lecons de g om trie el mentaire (2 volumes). Colin, Paris, 1898/1901. e e KLEIN, F. Elementarmathematik vom hoheren Standpunkte aus. (Teil 1 : Arithmetik ; Algebra ; Analysis / Teil 2 : Geometrie.) Teubner, Leipzig, 1908/1909. SCHUBRING, G. [2003] LEnseignement Math matique and the rst International Come mission (IMUK) : the emergence of international communication and cooperation. Ce volume, 4765.
LEARNING AND TEACHING OF ANALYSIS IN THE MID TWENTIETH CENTURY : A SEMI-PERSONAL OBSERVATION L tude et lenseignement de lanalyse e au milieu du vingti` me si` cle : une observation semi-personnelle e e par Man-Keung SIU
Cet article pr sente une observation semi-personnelle sur lenseignement de e lanalyse au milieu du 20e si` cle. Au lieu de compter uniquement sur les exp riences e e ` gl n es pendant ses ann es scolaires, lauteur rassemble aussi des informations a partir a e e darticles parus dans LEnseignement Math matique durant la d cennie 195565, de e e rapports pertinents sur les programmes scolaires et universitaires, de quelques manuels dun usage commun au milieu du 20e si` cle, et de correspondances et interviews e effectu es par lauteur avec des math maticiens de diff rents pays, qui etaient soit e e e etudiants soit jeunes professeurs dans les ann es 50 ou au d but des ann es 60. e e e ` e Cet article commence par d crire les math matiques a l cole pendant les ann es e e e 50 et le cadre de connaissances des lyc ens qui se lancaient dans l tude de lanalyse, e e mati` re habituellement d nomm e calcul innit simal dans la plupart des pays. Le calcul e e e e e diff rentiel et int gral a et introduit au lyc e au d but du 20e si` cle. Son programme e e e e e sest plus ou moins stabilis avant les ann es 50 en un enseignement g n ralement e e e e intuitif et d nu de rigueur math matique, mettant laccent sur les calculs m caniques e e e e plut t que sur la compr hension conceptuelle. En se basant sur sa propre exp rience, o e e ` e lauteur d crit le plaisir quil avait a r soudre des probl` mes de calcul innit simal e e e sans comprendre vraiment les th ories sous-jacentes. Ainsi, bien que le Th or` me e e e Fondamental de lAnalyse f t expliqu dans le manuel, lauteur n tait pas conscient u e e ` de limportance de ce beau r sultat a cette epoque-l` . Sans avoir saisi le concept dune e a int grale d nie comme limite dune certaine somme, il lui etait m me difcile dutiliser e e e le calcul int gral pour r soudre autre chose que des probl` mes st r otyp s. Malgr e e e ee e e cette compr hension impr cise (mais avec une comp tence technique raisonnablement e e e ` e sufsante), lauteur etait fascin par le calcul diff rentiel et int gral a l cole secondaire. e e e
180
M.-K. SIU
La discussion se poursuit sur la situation de lanalyse dans le mouvement des math matiques modernes. Lintroduction de sujets non traditionnels et linsistance e sur labstraction et les structures dans ce courant de pens e ne semblent pas avoir e affect fondamentalement lenseignement de lanalyse en tant que discipline scolaire, e ce qui signie quun difcile probl` me p dagogique restait difcile pendant ce moue e vement, et aussi apr` s. Il est vrai que, durant cete p riode, lanalyse etait enseign e e e e ` e a des el` ves de plus en plus jeunes en classes de niveau de plus en plus bas, que lemploi densembles et dapplications etait adopt pour la poursuite de lexactitude e ` math matique, et quil existait une tendance a labstraction. Mais, m me sans lemploi e e e densembles ni dapplications, m me sans la forte dose dabstraction, les el` ves e trouvaient lanalyse difcile parce quils ne comprenaient pas les concepts mis en jeu et donc ne pouvaient pas saisir la signication de cette mati` re. e ` La transition de l cole a luniversit posait (et pose encore aujourdhui) des e e probl` mes m me plus graves. Par exemple : e e a) Dans quelle mesure une transition en douceur est-elle en rapport avec une formation classique en g om trie ? Non seulement la connaissance de la g om trie e e e e ` e ` ee a l cole dautrefois permettait a l l` ve de shabituer avant ses etudes universitaires ` a la notion de d monstration et de logique, mais la g om trie en elle-m me est aussi e e e e ` ` une mati` re avec laquelle on peut sexercer a la discipline logique et en m me temps a e e ` ee limagination libre. D velopper une sympathie pour la g om trie permet aussi a l l` ve e e e de consid rer des probl` mes dans dautres domaines avec un point de vue g om trique. e e e e b) Dans quelle mesure le mouvement bourbakiste a-t-il inuenc lenseignement e universitaire de lanalyse, bien quil semble navoir eu que peu dinuence sur les ` e math matiques a l cole ? e ` e ` En ce qui concerne lenseignement de lanalyse a l cole ou a luniversit du e 21e si` cle, que peut-on apprendre en faisant r f rence au pass ? Les questions e ee e ` e contemporaines relatives a l tude de lanalyse, comme lemploi de la technologie moderne, la pertinence de cette mati` re dans la vie quotidienne, ou son rapport avec e ` la technologie informatique, n taient pas a lordre du jour au milieu du 20e si` cle. e e Cependant, les difcult s rencontr es par l l` ve a cette epoque-l` sont toujours les e e ee ` a ` difcult s rencontr es par l l` ve daujourdhui. Ce qui est a lorigine de ces difcult s, e e ee e cest la controverse perp tuelle entre les techniques et la compr hension, entre le concret e e et labstraction. Cette controverse, qui malheureusement se d veloppe parfois en une e ` fausse dichotomie st rile, etait un th` me retentissant au si` cle dernier a partir de e e e lintroduction du calcul innit simal dans les programmes scolaires jusqu` la nouvelle e a r forme qui a eu lieu dans les ann es 80 et 90. e e Lauteur conclut par une remarque sur lemploi du terme analyse au cours de lhistoire : on peut y voir comme un rappel de lunit fondamentale des quatre mati` res e e principales : analyse, arithm tique, alg` bre et g om trie, unit que, nous lesp rons, e e e e e e e nos el` ves peuvent comprendre et appr cier. e
LEARNING AND TEACHING OF ANALYSIS IN THE MID TWENTIETH CENTURY : A SEMI-PERSONAL OBSERVATION by Man-Keung SIU
1. A (SOMEWHAT
PERSONAL) PROLOGUE
When I was approached by the organizers of the Symposium as a prospective speaker, I certainly felt honoured but at the same time surprised and timorous. As students in mathematics it is true that we all study analysis ; and as teachers in mathematics we have all taught some courses in calculus or analysis. But I tend to view myself just as a plain mathematics teacher who believes in the value of the cultural and historical dimensions of this discipline in general education. So, this paper presents a semi-personal observation of the learning and teaching of analysis in the mid 20th century. Besides relying on the learning experience of my school days in the 1960s, I also gathered information from papers published in LEnseignement Math matique during the decade e 195565, from relevant reports 1 ) on the school and university mathematics curriculum, from some textbooks in common use in the mid 20 th century, and from correspondence or interviews with some mathematicians of different nationalities who were either students or young teachers in the 1950s and early 1960s. It must be admitted that this part of the project is not carried out systematically nor scientically and tends to be anecdotal. But it has a certain interest because it gives a non-Western perspective. Papers in LEnseignement Math matique usually recount learning and teaching in European countries, e
1 ) The following are particularly pertinent : [ICMI/UNESCO 196779], [UNESCO 1985], [Artigue 1991; 1996], Proceedings of International Congresses of Mathematicians (ICMs) and International Congresses on Mathematical Education (ICMEs) held between 1950 and 1980.
182
M.-K. SIU
sometimes also in North American countries, but very seldom in Asia. When I was a school pupil, the tidal wave of the new math movement had not yet swept over Hong Kong, so I was brought up in a more traditional curriculum, modelled after the British system, sometimes with a time lag, as Hong Kong was in those days a British colony. I began undergraduate study in 1963. It was in the university that I rst felt the impact of modern mathematics.
2. LEARNING
AND TEACHING OF ANALYSIS IN THE
1950S
AND
1960S
What was school mathematics like in the 1950s ? There is no dearth of reports on this question even if we conne our search to LEnseignement Math matique. Instead of going over them one by one, let me just quote a e passage from a report by Howard F. Fehr :
This, then is the picture of what the pupil has been taught. What does he really know ? This is hard to tell, but it can be said that the 15 year old in all countries, who has continued his study of mathematics through the rst 9 or 10 years of school can compute in a mature manner with the positive rational numbers, in a decimal system of notation, even though he cannot rationalize what he does ; he has a fairly useful and practical knowledge of geometry with respect to mensuration and common relationships ; and he can manipulate algebraic expressions and solve equations and problems in a structureless system of algebra. He can make simple deductions, but his entire concept of proof, if any, is limited to that of theorems in geometry. He really does not know what mathematics is, or how it is applied, but he has a large body of information, upon which, if he is inclined or interested, a study of mathematics can be built in the ages 16 years to 21 years. [H. F. Fehr 1959, 6667]
(As an aside, from what I observe of my own students today, most of them still do not know what mathematics is, but they have a far more meagre body of information mastered ; so, they experience difculty in continuing their study of mathematics.) Fehr continues with the aims of mathematical instruction labelled as : 1) mathematics for the better life, i.e. for its intrinsic value, or for its own sake ; and 2) mathematics for a better living, i.e. for its application to science, technology, and social problems that will result in more efcient practical day by day living. This should ring equally true forty years after it was written, but today people seem to pay more and more attention to 2), rather than 1), at most paying some lip-service to the latter. It was with the academic background depicted above that students in the upper secondary school embarked on the study of analysis, which was usually just called calculus in most places (as it still is today). Calculus had been
ANALYSIS IN THE MID TWENTIETH CENTURY
183
introduced into the upper secondary school curriculum in the early part of the 20th century. It met with some success, for instance in the reform of 1902 in France [Artigue 1996]. By the 1950s the syllabus was more or less stabilized to include : differentiation and integration ; simple applications such as rate of change, maxima and minima, area and volume, centre of mass, moment of inertia ; the trigonometric, logarithmic and exponential functions ; Taylor series expansions of functions. The instruction was in large part intuitive and informal, emphasizing calculation rather than conceptual understanding. For instance, the idea of limit was explained through an intuitive sense, that the dependent variable approaches a certain value as the independent variable approaches a given one. The derivative of a function f (x) at x a is described 0 at geometrically as the slope of the tangent line to the curve y f (x) the point (a f (a)) of the Euclidean plane R2 . From this description students were led to compute the limit of a Newton quotient [ f (a h) f (a)] h as h approaches 0 , from which point on they were drilled in the calculation of the derivatives of many different functions. The subject was meant to be preparatory for those who would go on to university and need it as a tool or prerequisite for further mathematical pursuits. Let me illustrate with my own learning experience. I came into contact with calculus before formal upper secondary schooling when in the preceding summer vacation my mathematics teacher kindly offered to give me extra lessons out of the popular American textbook Elements of the Differential and Integral Calculus by W. A. Granville [1904 (1929)]. One would read denitions like :
the variable is said to approach the constant as a limit when the successive values of are such that the numerical value of the difference ultimately becomes and remains less than any preassigned positive number, however small,

and recipes like :

GENERAL RULE FOR DIFFERENTIATION :
FIRST STEP. In the function replace x by x of the function, y y .
x , and calculate the new value
SECOND STEP. Subtract the given value of the function from the new value and thus nd y (the increment of the function). THIRD STEP. Divide the remainder y (the increment of the function) by x (the increment of the independent variable). FOURTH STEP. Find the limit of this quotient when x (the increment of the independent variable) varies and approaches zero as a limit. This is the derivative required.
184
M.-K. SIU
Integration is introduced as the inverse operation to differentiation, viz. to nd a function f (x) whose derivative f (x) (x) is given. The denite integral follows next, again with a recipe :
FIRST STEP. Find the indenite integral of the given differential expression. SECOND STEP. Substitute in this indenite integral rst the upper limit and then the lower limit for the variable, and subtract the last result from the rst.
In my school days I enjoyed doing all these, but I would not claim that I really understood what was going on. I still did not really understand what was going on even after studying later upper secondary school textbooks such as Techniques of Mathematical Analysis by C. J. Tranter [1957]. I solved quite a number of exercises from that book, most of them of a rather technical nature, such as :

If a b are positive and a
b a
1 , show that ab
1 a
2
Occasionally I ran into some theoretical discussion, such as :
If f (a) f (b) , there is at least one point x in the open interval (a b) at which f (x) 0 . If f (x) 0 for every x in the open interval (a b) then f (x) is strictly increasing throughout the interval.

Somewhere in the book there is a passage on the Fundamental Theorem of Calculus. But I was not aware of the signicance of this beautiful result at the time. In fact, without grasping the concept of a denite integral as the limit of a certain summation, I would have had difculty in applying integration at will to solve problems other than the stereotyped ones. If I had studied with all my might the book A Course of Pure Mathematics by G. H. Hardy [1908
If
cos(x sin ) d , show that d2 y dx2 1 dy x dx y 0

(n
2)(n
3)
dn y dxn
n(n
1)
dn 2 y dxn 2
Deduce that, when x
0, 0
Prove that, if y
x2 cos x , then x2 d2 y dx2 4x dy dx (x2 6) y 0
1 b
1 and deduce that 4 25 2
185
(1952)], I could have understood better. But the book was clearly beyond my comprehension at the time I bought the book because of its misleading title, since the subject I registered for the university entrance examination was called Pure Mathematics (to be distinguished from another subject called Applied Mathematics) ! However, one thing I remember vividly about reading that book is its appendix on two proofs of the Fundamental Theorem of Algebra with the beginning remark that it belongs more properly to analysis in the mind of a 16 year-old, it is strange to learn that the root of an algebraic equation has to do with innitesimal analysis. Despite this shadowy understanding (but with reasonably adequate technical competence in computation) I was thrilled with the subject of calculus. During the summer vacation when my teacher gave me the extra tutoring, I could solve a problem, albeit in a rather formal symbol-pushing manner without knowing what a differential equation is, that asked for the escape speed of a rocket. A rocket (of mass m ) is to be launched straight up from the surface of Earth (of mass M and radius R ). The student is asked to show that the minimum speed 0 at which the rocket must be launched to reach a distance
that the escape speed is
2GM . I started the calculation with the formula R
GMm (this much is physics) followed by a formal manipulation r2 d d dr d dt dr dt dr d GMm GM that yielded m d , or dr , so that dr r2 r2 m d dt
0
0
r0 R
GM dr ; r2
2GM wrong !) Finally, by putting r0 , I obtained the escape speed 0 R . This little feat on my part was particularly thrilling at the time when a little over three years before the rst Soviet sputnik had been launched into orbit, heralding the age of space travel !
r0

hence
GM r0 1 2 1 1 , or 2 0 GM r10 R , i.e. 0 2GM R r10 . 2 0 r R (Just to indicate how shadowy my understanding was at that stage, I should confess that I added the minus sign in the starting formula after I found out 1 1 1 2 that I arrived at an expression 2 0 GM 0 , which is blatantly
2 0
r0 from the centre of Earth is given by
2GM
1 R
1 r0
and to deduce
186 3. THE NEW
M.-K. SIU
MATH MOVEMENT AND ANALYSIS
The word sputnik, mentioned in a paper on mathematics education, is evocative of the new math movement. There is no need to go into the history of this movement, nor into the dispute on the pros and cons during and after it, although this is eventful, instructive and worth the discussion. There is a large body of literature on that. For an overview in the UK and the USA one can consult [Thwaites 1972] and [CBMS 1975]. Two thought-provoking papers published in LEnseignement Math matique discussed the issues in the e early phase of the movement : [Freudenthal 1963] and [Wittenberg 1965]. The introduction of untraditional subject matter and the emphasis on abstraction and structure in the new math movement did not seem to have a profound inuence on the learning and teaching of analysis as a school subject, in the sense that what was a difcult pedagogical problem before the movement remained a difcult pedagogical problem during and after it. It is true that in the new math movement, the subject of analysis, or at least notions and topics closely related to it (such as function, absolute value, error estimation, sequence and series, manipulation of inequalities, metrics, etc.), was taught in even lower grades for even younger pupils. Also, the language of sets and mappings was adopted for the pursuit of preciseness and there was a trend towards abstraction. But even without the language of sets and mappings, even without the strengthened dose of abstraction, students found analysis (calculus) difcult, not really because they could not do the (routine) calculations but because they did not understand and could not make good sense of the subject. The transition from school to university posed (as it still does today) even more serious problems. Several papers which appeared in LEnseignement Math matique in the mid 1950s were devoted to that issue [Freudenthal 1956 ; e Maxwell 1956 ; Behnke 1957 ; Delessert 1965]. Kay Piene also discussed this issue in an address in the ICM54 at Amsterdam [Piene 1956]. Again, let me illustrate with my own learning experience. Besides studying textbooks such as [Courant 1937], [Apostol 1957], or [Rudin 1953 (1964)], I had also access to Chinese textbooks which were modelled after the Soviet tradition of textbooks, written by authors like A. I. Khinchin 2 ) or G. M. Fichtenholz, that were popular in the USSR and in China in the mid 20th century. The course would start with a detailed discussion of the completeness of the real number system in its various formulations and the basic theorems about a continuous
2 ) A particularly lucid account, still worth studying today, is the little book Eight Lectures on Mathematical Analysis [Khinchin 1943 (1965)].
187
function dened on a closed interval. Thus I had the opportunity to sample the very rigorous treatment of analysis by the Soviet school, which was hard work but good solid training. In Europe, budding mathematicians in their rst year at university went through a similar rigorous diet of textbooks by authors like Edouard Goursat, Georges Valiron, or Jean Bass. The consequence of going through such a rigorous diet is that one either makes it or one gets irretrievably lost. In the mid 20th century, when tertiary education in most places still catered for elitism, this state of affair was allowed to go on. With the opening up of tertiary education in later decades, the problem is getting more and more noticeable and has to be faced and resolved.
4. TWO
QUESTIONS ON TRANSITION FROM SCHOOL TO
UNIVERSITY
There are two questions I wish to raise concerning the transition from school to university pertaining to the learning and teaching of analysis. a) By observing students I have taught from the mid 1970s till now, I nd that I was quite fortunate to have experienced a smoother transition. To what extent has this to do with my classical education in geometry ? As Richard Courant puts it in an introductory remark to his famous textbook on calculus [1937], its intimate association with geometrical ideas and its stress on individual niceties give the older mathematics a charm of its own. I was brought up with a large dose of synthetic geometry replete with lots of proofs and construction problems. Not only was I accustomed to the notion of proof and logic before starting undergraduate study, but in school geometry I tasted the joy of discovery and the joy of succeeding in understanding something which was tangible (you can at least draw some pictures even if you do not know why it has to be like that at rst) but not obvious (you do not know why it is like that at rst). Geometry is a subject in which one can exercise logical discipline and free imagination at the same time. Developing a liking for geometry also enabled me to look at problems in other subjects from a geometric viewpoint. This helped in particular in the study of analysis. After all, calculus is the process of linear approximation, and linear problems fall within the purview of linear algebra, which is in itself akin to geometry. Of course, it does not work all the time and different persons are accustomed to different ways of thinking ; nevertheless it offers an alternative. Many students today are not accustomed to this exibility in framework in their study of mathematics.
188
M.-K. SIU
b) How extensive is the inuence of Bourbakism felt on the teaching of analysis in the university ? In school mathematics, in my experience, there was relatively little inuence. I rst learnt about the work of Bourbaki from my teachers at university. I cannot say much more further on this issue as I have not found out enough, and besides it would be somewhat remote from school mathematics, which is what we are more focused on in our discussion.
5. WHAT
CAN WE LEARN FROM LOOKING AT THE PAST ?
What can we learn from looking at those bygone days as far as the learning and teaching of analysis in the school or university classroom of the 21st century are concerned ? Back in those days, the all-purpose electronic computer was just making its d but ; the hand-held calculator was still a luxury e in the classroom ; the Internet was not thought of even in works of science ction ; the mathematics curriculum was not as broad and as diversied ; the application of analysis was mainly conned to physics and engineering. As a result, contemporary issues in the learning and teaching of analysis, such as the use of modern technology in the classroom, the relevance of the subject in daily life or the relationship of the subject to information technology, were not yet on the agenda. However, the difculties a student encountered in those days are still the difculties a student encounters today. In this sense, many of these contemporary issues, though they may play a signicant role in improving the learning and teaching of the subject, are secondary since they may well breathe new life into the subject but they are not at the root of the difculty. What is at the root is the perennial controversy between computational skill and understanding, between concreteness and abstraction. This controversy, which can sometimes unfortunately develop into an unnecessary false dichotomy, is a theme which reverberated throughout the last century from the introduction of calculus into the school curriculum to the calculus reform of the 1980s and 1990s. In this connection let me quote two relevant passages from two great teachers :
The point of view of school mathematics tempts one to linger over details and to lose ones grasp of general relationships and systematic methods. On the other hand, in the higher point of view there lurks the opposite danger of getting out of touch with concrete details, so that one is left helpless when faced with the simplest cases of individual difculty, because in the world of general ideas one has forgotten how to come to grips with the concrete. The reader must nd his own way of meeting this dilemma. In this he can only
189
succeed by repeatedly thinking out particular cases for himself and acquiring a rm grasp of the application of general principles in particular cases ; herein lies the chief task of anyone who wishes to pursue the study of Science. [Courant 1937, 23] Abstraction was not a hormone which can be imposed from outside, but one that the patient must generate for himself in response to appropriate stimulation. [Quadling 1985, 94]
Finally, let me inject a historical remark on the term analysis. In the ancient Greek usage of this word it means, in contrast to synthesis, the process of working backward from what is sought until something already known is arrived at. Towards the end of the 16th century Francois Vi` te e used the term analysis to denote algebra, since he did not favour the word algebra (coming from the Arabic word al-jabr) which has no meaning in any European language. In his book In artem analyticen isagoge, much of the algebra developed is motivated by the intention to solve geometric problems. In this sense it is related to the ancient usage of the word in describing the process. This was brought even more into focus by the work of Ren e Descartes, as illustrated by the famous appendix La g om trie. e e When calculus came on stage in the era of Isaac Newton and Gottfried Wilhelm Leibniz, they regarded the subject as an extension of the algebra of the innite, in which lots of functions were expressed as power series that behave like polynomials, just longer ! This was again emphasized in the work of Joseph Louis Lagrange with the title Th orie des fonctions analytiques. e Leonhard Euler titled his book on calculus Introductio in analysin innitorum. By and by the term analysis was used to denote the study of calculus and its extension. In the preface to an unpublished book on analysis, Henri Lebesgue discusses the relationship and distinction between arithmetic, algebra and analysis (published posthumously as [Lebesgue 1956]). Gustave Choquet [1962] emphasizes the inseparable relationship between algebra, geometry and analysis throughout his paper. Thus, the subject of analysis is closely tied in with the subjects of arithmetic, algebra and geometry. Maybe this historical episode is a reminder of the integrated unity of the four basic subjects in the mathematics curriculum, which we wish our students to realize and to appreciate.
REFERENCES APOSTOL, T. Mathematical Analysis : A Modern Approach to Advanced Calculus. Addison-Wesley, Reading (Mass.), 1957. ARTIGUE, M. Analysis. In : Advanced Mathematical Thinking, ed. D. Tall, 167198. Kluwer, 1991.
190
M.-K. SIU
ARTIGUE, M. R formes et contre-r formes dans lenseignement de lanalyse en lyc e e e e (19021994). In : BELHOSTE, B., H. GISPERT & N. HULIN (eds.) Les sciences au lyc e, un si` cle de r formes de lenseignement des math matiques et de la e e e e ` e physique en France et a l tranger, 192217. Paris, VuibertINRP, 1996. BEHNKE, H. La tension entre lenseignement secondaire et lenseignement universitaire en Allemagne. LEnseign. Math. (2) 3 (1957), 237250. CBMS. NACOME Report On Overview And Analysis Of School Mathematics Grade K-12. Reston, NCTM, 1975. CHOQUET, G. Lanalyse et Bourbaki. LEnseign. Math. (2) 8 (1962), 109135. COURANT, R. Differential and Integral Calculus. Volumes I and II. 2nd edition, translated by E. J. McShane. London, Blackie, 1937. ` DELESSERT, A. Quattend de luniversit le matre enseignant les math matiques a e e l cole secondaire ? LEnseign. Math. (2) 11 (1965), 309320. e FEHR, H. F. The mathematics education of youth : a comparative study. LEnseign. Math. (2) 5 (1959), 6178. FREUDENTHAL, H. Relations entre lenseignement secondaire et lenseignement univer` sitaire en Hollande. (Conf rence faite a Gen` ve au Symposium en lhonneur de e e Henri Fehr). LEnseign. Math. (2) 2 (1956), 238249. Enseignement des math matiques modernes ou enseignement moderne des e math matiques ? LEnseign. Math. (2) 9 (1963), 2844. e GRANVILLE, W. A. Elements of the Differential and Integral Calculus (revised by P. F. Smith and W. R. Longley, 1929). Ginn, Boston, 1904. HARDY, G. H. A Course of Pure Mathematics. (Tenth edition 1952 ; originally published in 1908.) Cambridge University Press, 1908. ICMI/UNESCO. Tendances nouvelles de lenseignement des math matiques. Vol. IIV. e ICMI/UNESCO, Paris, 1967, 1967, 1973, 1979. KHINCHIN, A. I. Eight Lectures on Mathematical Analysis. Heath, Boston, 1965. (First published in Russian in 1943.) ` ` LEBESGUE, H. De larithm tique a lalg` bre et a lanalyse math matique. LEnseign. e e e Math. (2) 2 (1956), 4960. ` MAXWELL, E.-A. From secondary school to University. (Conf rence faite a Gen` ve e e au Symposium en lhonneur de Henri Fehr). LEnseign. Math. (2) 2 (1956), 307313. PIENE, K. School mathematics for universities and for life. In : Proceedings of International Congress of Mathematicians 1954, Amsterdam, Volume III, 318324. Amsterdam, North Holland, 1956. QUADLING, D. Algebra, analysis, geometry. In : [UNESCO 1985], 7996. RUDIN, W. Principles of Mathematical Analysis. (2nd edition 1964.) McGraw-Hill, New York, 1953. THWAITES, B. (ed.) The School Mathematics Project : The First Ten Years. Cambridge University Press, 1972. TRANTER, C. J. Techniques of Mathematical Analysis. The English Universities Press, 1957. UNESCO. Studies in Mathematics Education, Vol. 4 : The Education of Secondary School Teachers of Mathematics, ed. R. Morris. UNESCO, Paris, 1985. WITTENBERG, A. Priorities and responsibilities in the reform of mathematical education : An essay in educational metatheory. LEnseign. Math. (2) 11 (1965), 287308.
ANALYSIS 2000 : CHALLENGES AND OPPORTUNITIES Lanalyse en lan 2000 : d s et opportunit s e e par Lynn Arthur STEEN
Dans cette contribution on essaie de tracer les lignes de d veloppements possibles de e lenseignement de lanalyse. Il importe de souligner que dans le monde de l ducation e le mot analyse est plut t utilis au sens de lam ricain calculus. Il sagit de calcul o e e diff rentiel et int gral concu pour les masses ; il se concentre sur des proc d s et des e e e e recettes, alors que lanalyse sint resse aux preuves et aux d nitions fondamentales. e e ` e En tout cas, lanalyse noccupe plus a pr sent la position dominante qui etait la ` sienne il y a cinquante ans. Quels sont les facteurs contemporains de nature a inuencer le futur de lenseignement de lanalyse ? Dabord on enum` re des facteurs li s a lenvironnement : facteurs sociaux (respone e ` ` sabilit publique, mutations technologiques, enseignement a distance, economie gloe e e bale) ; facteurs li s a l ducation (m thodes d valuation, nombre elev d tudiants, e ` e e e dont la pr paration nest pas homog` ne), facteurs math matiques (diversication des e e e math matiques, math matiques pour tous). e e A propos des r formes de lenseignement du calcul diff rentiel et int gral, on e e e mentionne deux mouvements qui se sont d velopp s aux Etats-Unis. Il y a vingt ans e e lint ret pour linformatique a remis en question la supr matie de lanalyse comme e e porte dentr e aux math matiques universitaires. Ce mouvement a eu le soutien de e e scientiques, dadministrateurs, de politiciens et de math maticiens. Dans le m me e e temps, mais ind pendamment, le cours intitul Calculus devenait un symbole de e e e haute qualit de linstruction : les parents comme les el` ves revendiquaient pour ce e cours un r le central dans le programme du secondaire. Le but n tait nullement de o e perfectionner ce cours, mais de permettre au plus grand nombre de passer lexamen, e si possible avec des notes elev es. La r forme a chang les cours danalyse plus dans la p dagogie et le contexte e e e que dans les contenus. On met laccent surtout sur les ordinateurs, les calculatrices de poche, le travail en groupes, les projets d tudiants, l criture, la mod lisation, e e e l tablissement de liens et les applications. On demande aux etudiants davoir une plus e
192
L. A. STEEN
profonde compr hension des relations entre les formules, les graphes, les nombres et les e descriptions verbales. Mais limpact le plus important de la r forme de lenseignement e e de lanalyse a et lengagement des professeurs duniversit dans la discussion sur e e lapprentissage des math matiques el mentaires destin es au plus nombre. e e Lavenir de lenseignement du calcul diff rentiel et int gral met en jeu plusieurs e e questions. Dabord il y a la mani` re dont les etudiants apprennent le sujet et les obstacles e ` cognitifs. Par exemple, il faut evoquer le passage de lalg` bre a lanalyse, linuence du e contexte et des exp riences quotidiennes sur lapprentissage, le rapport entre l ducation e e secondaire et tertiaire, le probl` me de la rigueur. Mais il est egalement important de e se demander comment les mutations technologiques ont transform lapproche de e ` e lanalyse. Comme les probl` mes dapprentissage sont difciles a r soudre, peut- tre e e les math maticiens devraient-ils prendre connaissance des recherches qui se font dans e les neurosciences sur les constructions mentales des math matiques. e
ANALYSIS 2000 : CHALLENGES AND OPPORTUNITIES by Lynn Arthur STEEN
Talking about challenges facing analysis education at the turn of the new century is a daunting task, well beyond the wisdom of any one person. However, it does provide a rare opportunity to think about what might be rather than what is, to imagine inquiries unencumbered by the constraints of current literature or prevailing orthodoxy. Thus my role is not that of a mathematician who proves theorems that last forever, but that of a prophet who imagines a future that is yet to emerge. I take some comfort in the knowledge that similar predictions on other anniversary occasions have set a very low standard for this genre. Instead of beginning as a mathematician would, with denitions and notation, I shall begin, as an amateur prophet might, with clarications and distinctions. Although my assigned subject is analysis, in the world of education analysis tends to mean calculus. Indeed, the vast majority of analysis enrollments are in elementary calculus. The distinctions between what mathematicians call analysis and what students call calculus are clear to anyone who has taught both courses. Calculus focuses on procedures and templates, analysis on denitions and proofs. Calculus is for the masses, analysis is for the mathes those who plan to specialize in mathematics. In this paper, my focus will be on calculus. A second distinction is about mathematics vs. education. During the last half-century mathematics has expanded enormously, both in the diversity of its specialties and in the pervasiveness of their roles in society. These socalled mathematical sciences encompass a diverse and rapidly expanding part of human intellectual accomplishment. Although still vigorous both in its own right and as a supporting tool for other parts of mathematics, analysis now represents a much smaller fraction of mathematical practice than it did fty years ago, before linear programming or digital computers, before combinatorics or bioinformatics, before data mining and string theory.
194
L. A. STEEN
Paradoxically, during this same half-century the role played by calculus in education has expanded enormously, often irrationally. Today the mathematical focus of secondary school is calculus, not mathematics. Even as the mathematical sciences have built avenues of intellectual exchange extending in many different directions, the signposts of society still direct everyone to enter mathematics on the traditional highway of calculus. The question we must confront at the beginning of this new century is whether this traditional route still serves mathematics well or whether it might not be prudent to explore other options through which students can enter the world of mathematics.
1. ENVIRONMENTAL
SCAN
As is fashionable these days among those whom we in the United States suspiciously call policy wonks, I begin with an environmental scan to survey briey some powerful contemporary factors that are likely to have signicant bearing on the future of calculus. First are four societal factors :
Public accountability. All over the world, government leaders are raising their expectation that universities serve broad public purposes, not merely elite professional interests. This demand arises both from increasing democratization of governments who must now be at least somewhat responsive to the people they govern and from rapid increases in secondary and tertiary enrollments that are supported to a great degree by public funds. Accountability pressures on mathematics departments are now external, not just internal ; they arise as often from parents and politicians as from professionals and peers. Technology. As computers become standard tools of employment and research, prociency in professional use of technology becomes an obligation of education. Mathematics educators now must deal not only with the question of how technology can enhance (or impede) mathematical learning and how technology changes priorities for mathematics content, but also how technology changes the way mathematics is expected to be performed. Technologys challenge to mathematics education is now about ends as well as means. Telecommuting. Mathematical skills, being purely cerebral, can be bought and sold anywhere on a worldwide market that is increasingly linked with high speed Internet connections. Thus the market for mathematical skills is no longer local or parochial but international. As graduates can now
ANALYSIS 2000 : CHALLENGES AND OPPORTUNITIES
195
sell technical skills to employers anywhere in the world, so students and teachers can also join a community of learners without borders. Thus despite lingering national differences in mathematics curricula, calculus as a subject is increasingly shaped by international contexts.
Global economy. As nations compete in a technologically sophisticated international economy, the demand for technically trained employees who can manage complex industries is increasing far more rapidly than is the demand for researchers and scientists. Thus the economic demand for mathematical skills is now tilted in the direction of breadth rather than depth, for the practical skills of calculus rather than the theoretical skills of analysis, and even more for the data-based skills of statistics rather than the function-based skills of algebra.
Now three educational factors :
Diversication of education. With the growth of a parallel universe [Adelman 2000] of higher education on-line courses, for-prot colleges, certicate programs, and virtual universities students now have many more options for their education. For better or worse, school and college mathematics departments will no longer have hegemony over calculus. Assessment. More and more, educational quality is being measured by outputs (student learning) rather than inputs (faculty teaching). Increasingly, calculus will be judged not by the syllabus, textbook, or instruction, but by the degree to which students can demonstrate that they meet the objectives of the course many of which are totally absent from traditional tests. Enrollment pressure. In virtually every country the number of students completing secondary school and entering post-secondary education is rising, bringing to upper secondary and university levels students of very different skills, backgrounds, and motivations. These changes, worldwide in scope, increase enormously the pedagogical challenges of teaching calculus.
Finally, two mathematical factors :
Diversication of mathematics. The expansion of mathematical methods into such diverse areas as genetics, nance, and even cinema has signicantly changed the balance of mathematical practice. New tools developed for these new applications, especially in combinatorics, geometry, and data analysis, have displaced analysis from the leading role it has played for
196
L. A. STEEN
nearly three hundred years during which it has been at the center of the mathematical universe. As analysis no longer plays the lead role in mathematics, so calculus is gradually losing its claim to be the centerpiece of mathematics education.
Math for all. Worldwide, educational and governmental leaders now include mathematics as part of the common core of learning expected of all students. However, especially in light of the changes in the mathematical sciences, one must now ask whether math for all should continue to mean, as it has in recent decades, calculus for all. There is some evidence that accelerated and excessively narrow mathematical requirements have created a backlash, leading some countries to scale back mathematical requirements in the schools.
2. CALCULUS
REFORM
In the United States, and to varying degrees in other countries, one of the major factors inuencing calculus in recent years has been a movement marching under the banner of calculus reform. The progress and setbacks of this campaign provide a valuable case study in change, illustrating both how calculus responds to external pressures as well as how it manages to retain surprising equilibrium despite these pressures. I apologize for focusing in this international forum on a movement that arose and remains anchored in the United States. I do this for both weak and strong reasons. The weak reason is that it is what I know best. The strong reason is that students and teachers in the United States come literally from all over the world. Thus the challenges we face are in some ways representative of the challenges faced by students and teachers in other countries. Indeed, our mixture of cultures in a single nation, a single city, a single school or university, even a single classroom, creates extraordinary instructional challenges that may be as internationally representative as anywhere on earth. Twenty years ago, rising interest in computer science began to challenge the primacy of calculus as the gateway to university mathematics. Computing both pulled students away from calculus and challenged the importance of topics taught in calculus. One response was a campaign led by US college
197
and university mathematicians to make calculus more lean and lively, to make it a pump not a lter in the educational pipeline [Douglas 1986 ; Steen 1988]. The reform movement drew support from very different sources from scientists who were frustrated by the inability of students to use calculus intelligently in real applications, from administrators who were angered by high failure or withdrawal rates from calculus courses, from politicians who saw in the global economy an increased need for technicians rather than theoreticians, and from mathematicians who recognized that calculus instruction had become, in Bernard Hodgsons memorable image, lenseignement scl ros [Hodgson 2000]. e e Simultaneously but independently, calculus in the United States also became a political totem, a supposedly objective and unassailable surrogate for high standards to which politicians could appeal in supporting or attacking various education proposals. Backed by political rhetoric that proclaimed calculus as the epitome of academic accomplishment, parents and students began pushing calculus into the secondary curriculum, primarily through the vehicle of the Advanced Placement (AP) program. This political campaign had nothing to do with reform but everything to do with status and access. Its goal was not to improve calculus but to enable more secondary school students to pass calculus, preferably with high grades (see, e.g., [Stewart 1997] and [Spence 2000]). Needless to say, these two movements had rather different aims and objectives. The goal of the AP course soon became numbers : more offerings, more participants, more passing grades. The course itself is anchored by a traditional syllabus set by an external committee that also establishes the national exam for the course. Recently the syllabus was changed slightly to reect some aspects of the calculus reform movement, but even these small changes produced great anxiety among AP teachers who depend on the courses stability and predictability for their success in getting students through the exam. Even as AP calculus became the gold standard of secondary school mathematics, the tertiary level calculus reform movement tried with increasing energy to destabilize those entrenched aspects of postsecondary calculus that, in reformers eyes, were not serving students well. For better or for worse, reform calculus became a magnet for unfullled goals of every prior progressive movement in mathematics education. Ten years after calculus reform had begun, a federally sponsored assessment of the movement found over a dozen different goals, only a few of which had anything to do with the content of calculus [Tucker & Leitzel 1995].
198
L. A. STEEN
Three content goals, not among the rst on the list, did imply uency in traditional skills :

Develop improved in-depth understanding of specic mathematical concepts. Employ calculus techniques in other disciplines and in novel situations. Reason analytically, qualitatively, and quantitatively. Represent problems algorithmically, graphically, verbally, numerically, and symbolically. Use technology in solving problems. Deal with complex, often ill-dened problems. Translate problems from one form to another. Communicate mathematical ideas both in writing and orally. Become independent mathematical learners. Work effectively in groups. Understand the role of experimentation, conjecture, verication, and abstraction. Develop positive attitudes about ones ability to do mathematics successfully. Appreciate mathematics elegance and structure. Pursue further work in quantitative elds.
Several cognitive goals stressed robust problem solving skills :
Some behavioral goals concerned how students work and interact with others :
Finally, attitudinal goals focused on students feelings about mathematics :
3. EVALUATION
AND ASSESSMENT
It is now fteen years since the movement to reform calculus began, time enough to assess its short-term impact. It turns out that reform courses and textbooks in the United States the text largely denes the course do not differ much from traditional courses in content, but they do differ signicantly in pedagogy and context. Reform calculus typically places much greater emphasis on calculators, computers, cooperative learning, students projects, writing, modeling, connections, and applications ; it makes only modest changes in the way limits, derivatives, integrals, and series are approached or developed. Although evidence is scarce and unreliable, the consensus of experts and the conclu-
199
sions of several studies is that the majority of students in the United States still study calculus from texts and in classrooms that have changed very little during the fteen years of the calculus reform movement. However, a sizeable minority typical estimates range from 25% to 40% encounter calculus in courses that attempt to implement at least some of the strategies common to reform courses [Tucker & Leitzel 1995, 2327 ; Roberts 1996, 1578]. These students do face different expectations. As evidenced by nal examinations, students in fully reformed courses, compared with their peers in traditional courses, are expected to have much deeper understanding of the relations among formulas, graphs, numbers, and verbal descriptions. For example, in a traditional US course, a typical exam problem at the end of the rst term would give the formula for a function and ask the student to nd the derivative and then sketch both the function and its derivative. In a reform course, a typical problem might give students six or eight unlabeled graphs and ask them to decide which are derivatives of which others and to explain how they gured it out [Roberts 1996, 74132]. Convincing evidence of changes in students mathematical performance, cognitive habits, work behavior, or attitude towards mathematics is both scarce and decidedly mixed. Since the goals of a reform course are so different from those of a traditional course, the lack of clear-cut evidence is not surprising [Schoenfeld 1997 ; Gold, Keith & Marion 1999, 229256]. Different students learn different things well, and different things poorly ; some thrive under the reform regimen, others chafe. Even good students often encounter difculty when they shift from one style of course to another because it upsets their own expectations of what is required. Direct comparison of outcomes is difcult, if not nonsensical a bit like comparing the outcomes of a course in Shakespeare with a course in James Joyce. By far the greatest impact of the calculus reform movement is the engagement of university mathematicians, sometimes for the rst time, in serious, productive discussions about teaching and learning elementary massmarketed mathematics [Tucker & Leitzel 1995, 3642]. Ideas for improved pedagogy spin off from these discussions and take root in other courses across the curriculum. Many university mathematicians have discovered that they share a common agenda with secondary school mathematics teachers who are working to implement new standards for school mathematics. Perhaps even more surprising, many university administrators nd in the reform movement meritorious evidence of scholarship that both advances the art of teaching and resonates with institutional evaluation for tenure and promotion [Joint Policy Board 1994].
200
L. A. STEEN
4. INQUIRY
AND SCHOLARSHIP
Calculus reform is one manifestation of efforts to improve the teaching and learning of analysis and more broadly, of higher mathematics. Another is the growing interest in research in mathematics education at the upper secondary and undergraduate levels (where analysis education begins). These are not unrelated, of course, since most advocates of calculus reform claim to ground their recommendations in the results of research. (Not surprisingly, most critics of the reform movement (e.g., [Wu 1997 ; Askey 1999]) dispute the relevance of that same research.) I tend to adopt a broad view of research, one aptly expressed by the phrase disciplined inquiry that Mogens Niss employed in his plenary address at ICME-9 [Niss, forthcoming]. This phrase nicely subsumes four distinct yet interconnected aspects of scholarship (of discovery, of synthesis, of application, and of teaching) that Ernest Boyer identied in his well-known monograph Scholarship Reconsidered [Boyer 1990]. The literature is lled with evidence of more focused or more idiosyncratic denitions, especially of educational research. I am not an expert in this area, but neither are the tens of thousands who teach calculus. So in this respect, I stand in their shoes as one who is more a practitioner than a theoretician of education. It is not natural for me, or for them, to ask about belief systems or metacognition, much less about APOS (actionprocessobjectschema) stages or statistically signicant p -values (e.g., [Kaput & Dubinsky 1994 ; Dubinsky, Kaput & Schoenfeld 19941998]). Valuable as these may be to experts, they do not communicate in the language spoken by practitioners. So instead of attempting to outline a research agenda cast in the traditional language of educational research, I want to suggest a variety of issues concerning the future of calculus that I believe could benet in coming years from disciplined inquiry and reconsidered scholarship. Some of these issues are about cognition and understanding, some about teaching and learning, and some about policy and practice. Professionals may consider some of these issues out of the normal bounds of educational research. But I believe, and I hope you will too, that not only calculus but also analysis and, indeed, all of mathematics would nonetheless benet from these kinds of disciplined inquiry. Before delving into the issues, I need to make one more distinction that between descriptive and normative questions, between inquiry into what is vs. questions about what should be. For example, in the history of calculus reform, one of the key results from the what is inquiry was evidence of astonishingly high rates of dropout, failure, and repetition in calculus as taught
201
in US colleges and universities. No other subject was even close. Another indicator, perhaps a bit more subjective, was the widespread sense among both mathematicians and users of mathematics that even students who completed calculus with good grades were, in too many cases, incapable of using it intelligently and expeditiously. These failures of traditional calculus were documented by descriptive investigations. In contrast, reform calculus the movement was the result of disciplined inquiry into the normative question of what calculus should look like if we want to improve retention and improve performance in subsequent courses. As we have seen, this inquiry also led to additional goals, to cognitive, behavioral, and attitudinal expectations that were not central to the reform effort in its early, descriptive, hand-wringing stage. For some mathematics teachers, these newer goals for example, writing, using technology, mathematical modeling, and group projects are strategic objectives aimed at bringing about more traditional goals of improved retention and performance. For others these goals are ends in themselves, dimensions of what it means to learn calculus in the twenty-rst century. In most instances, descriptive and normative inquiries occur simultaneously on parallel paths. Ideas and evidence from one inuence the other, generally for the betterment of both. Nonetheless, insofar as possible, I believe that it is important to keep the distinction clear, much as journalists, ideally, attempt to maintain a relatively clear line between reporting and advocacy. That said, I begin with some issues that I believe merit disciplined descriptive inquiry.
5. DESCRIPTIVE
ISSUES
How many students study calculus ? What do they learn ? This is basic. In the United States, we gather some data addressing this issue, but it is relatively infrequent and not easy to interpret because much calculus in secondary schools is embedded in other courses, or taught supercially as a warm-up for later formal study [Loftsgaarden, Rung & Watkins 1997]. Other nations may do their own studies, but there seems to be no source of accurate international enrollment information on courses such as calculus that are optional and taught in many different kinds of educational settings. (And now we have the added complication of on-line Internet courses.) Calculus is one of relatively few optional courses of broad signicance that is taught everywhere in the world, so it may merit special attention as a surrogate for the quality and extent of advanced education in different nations.
202
L. A. STEEN
In what ways does calculus depend on its setting ? This inquiry probes the universality of calculus by seeking explicit information about variability that may be due to external contexts : different nations, different educational settings (secondary school, technical college, university), different student abilities, different settings for diverse applications (physics, biology, business). What is common and what is variable ? Are there any universals in terms of denitions, problems, theorems, applications, sequencing ? Are there particulars that are either widely adopted or widely ignored depending on context ? An international survey of calculus instruction could help dene the subject by identifying what is central and what is peripheral. How well do calculus tests assess the goals of calculus courses ? This is an important empirical question because tests, not syllabi, determine what most students actually learn. In traditional courses in the United States, tests ask primarily for routine calculations plus one or two predictable applications and perhaps some simple proofs [Steen 1988, 177211]. Exams for reform courses include a wider variety of conceptually interesting problems, often expecting uency not just with formulas and functions, but also with graphs, numerical tables, and verbal descriptions [Roberts 1996, 74132]. But in neither case does one nd many questions that reveal the broader goals of reform calculus nor the deeper goals of traditional courses. How important is calculus now that graphing and symbol manipulating software is widely available ? Is calculus still as important as it was during the three hundred years between Isaac Newton and Bill Gates when it was the only tool available for most scientic models ? Put another way, if digital computers had been invented before calculus, would calculus ever have gained the central position it now occupies in mathematics education ? Now that virtually any set of differential equations can be solved digitally, closedform analytical solutions are no longer the gold standard of mathematical modeling : visual representations of changes in behavior under different values of parameters offer far more insight than do symbolic solutions. This question calls for empirical investigation into the practice of mathematics to determine the degree to which the new digital empire has replaced the analysis empire that has been in power for most of the last three centuries. Results of this investigation will almost surely lead to a revised map of mathematics. Which aspects of calculus require human expertise and which are best performed by computers ? This inquiry into the content of calculus in the computer age naturally splits into two parts calculus as it is practiced and calculus as it is taught. Long experience has provided mathematics teachers
203
with a pretty good mental map of the logical and instructional sequencing appropriate for calculus in its traditional form when most student work is done with paper and pencil. But now that computer software can do much of what students have historically been asked to do, and now that computer systems are widely used in every profession where calculus plays a signicant role, we need to envision calculus differently. Technology is doing for calculus what a new subway does for a city and what international air transportation is doing for the globe : it is realigning relationships and changing psychological distances. A new map of calculus would reect these changed relationships. What mathematical uses of calculators and computers are inappropriate ? Twenty years experience shows rather convincingly that calculators can, if wisely used, enhance students experiences in mathematics class. In these masterful classes, students learn mathematics well, gain facility with a handheld mathematical machine, and emerge enjoying mathematics more than those who study the same material in more traditional contexts. However, university mathematicians universally complain that large numbers of students use calculators inappropriately either to perform calculations that they really should do in their heads or on paper, or to give approximate numerical answers 2 ) would reveal much greater to problems where an exact answer (e.g. understanding than the numerical approximation (4.4428829). Mathematicians are not alone in this concern ; older adults often complain when they see young store clerks rely on calculators to perform simple calculations. So this inquiry is not about whether calculators enhance or diminish learning, but whether it is possible to distinguish between appropriate and inappropriate uses of calculators in a way that will be useful for teaching and learning. (If this could be done, then perhaps mathematicians might be able to reach consensus, now regrettably lacking, on the appropriate role of calculators in calculus.) How does the way calculus is taught inuence students ability to use calculus in further study and professional practice ? Most research into pedagogy is concerned about its relation to learning. This proposed inquiry points in a different direction by asking not about the effect of teaching practice on student learning but about its effect on students subsequent ability to use what they have learned. Students notorious inability to transfer learning from one context to another from mathematics class to economics class, for example, or from classroom to work place is widely recognized as a generic problem of learning, not a special disability of mathematics. Student resistance to employing good writing outside of language classes is equally
204
L. A. STEEN
well known. The issue for inquiry is whether and how the circumstances of learning can facilitate transfer of what has been learned. How much of calculus is learned outside of calculus class ? We know from many studies that young children pick up considerable mathematics outside of school : at home, in the playground, in stores, on television, and now at the computer. Childrens mathematical minds are shaped substantially both by what happens in and outside of mathematics class. Historically, this has also been true for calculus students. Typically, students learned calculus deeply primarily by using it for physics. (And they learned algebra uently primarily by using it for calculus.) Today relatively few mathematics students study physics, but many study business and economics where calculus-based models abound. We need to learn systematically (rather than anecdotally) just how important these out-of-math-class experiences are for full mastery of calculus. Does it matter much for mastery of derivatives and integrals if the primary application is to physics the historic taproot of calculus or to newer sciences such as economics or biology ? More interestingly, can average students really learn calculus well if they study only calculus ? Might it be that the subtle nature of innite processes requires the anchor of a real model (rather than only mathematical denitions) for the mind to construct an appropriate and usable mental model of calculus ? What special cognitive hurdles are involved in the transition from algebra to analysis ? All mathematics teachers recognize an enormous gap between the knowledge and skills developed in secondary school courses in geometry and algebra knowledge that is primarily static and procedural and the dynamic subtlety of limits and limit-based concepts such as the real number line, the derivative, and the integral. This gap is partly due to conicts between ordinary and mathematical language (e.g., limit as a barrier ; continuous as incessant ; innite as unfathomable), but even more to the increasing level of abstraction that is inherent in nested quantication (e.g., for every epsilon there exists a delta such that ). It took mathematicians two centuries to gure out how to express the fundamental denitions of calculus in such a way as to avoid contradictory inferences when reasoning about innite processes. It is not enough for calculus teachers to just understand the logical resolution of these paradoxes of the innite the so-called arithmetization of analysis introduced in the nineteenth century. The issue requiring inquiry is how these logical resolutions may in fact exacerbate (rather than resolve) the psychological impediments students face in dealing with innite processes.
205
What characteristics of students, teachers, schools, and policy account for the signicant differences between secondary and tertiary education ? Calculus, the introduction to analysis, straddles the active fault line where secondary education pushes up against the plate of tertiary education. Stress is evident everywhere on students, teachers, schools and discontinuities abound. Students face the daunting challenge of leaving the security of algebra for the uncertainties of analysis at the same time as they traverse the steep terrain leading from secondary to higher education. Although anticipation of educational and intellectual hazards will not eliminate them, foreknowledge can help both students and teachers develop strategies to minimize their negative effects. The purpose of this inquiry is to identify and describe special impediments to students intellectual growth in these years due to factors other than the special character of analysis.
6. NORMATIVE
ISSUES
Shifting gears, I conclude with some normative questions, with suggestions for more speculative and subjective inquiry into possible futures for calculus, and ipso facto, for analysis education. What is the best way to introduce students to university mathematics ? In particular, is calculus the right course for the majority of students ? No one disputes that calculus is both signicant and sublime. But are these sufcient reasons to require that all students enter higher mathematics through this single gateway ? Students in the social, behavioral, and life sciences now the majority of clients of mathematics need a much broader portfolio of mathematical methods, notably statistical, combinatorial, and computational. Other themes, such as optimization or modeling, that represent more widely applicable areas of mathematics can provide intellectual challenges equal to those of calculus as well as a synthesis of methods from algebra, analysis, geometry, and combinatorics. Might multiple gateways to mathematics help reverse the worldwide decline in students who specialize in university-level mathematics ? Should mathematicians welcome increasing numbers of students to calculus ? Many of the complaints one hears from calculus teachers about students lack of preparation or motivation are consequences of societys pressure, wisely or not, to push more and more students into higher levels of mathematics. Many mathematicians talk as if they would much prefer that mathematics had remained an elite subject for a select few rather than become a basic subject for mass education. Analysis (that is, calculus) is where this argument
206
L. A. STEEN
is joined, since everyone acknowledges that all students must study the other foci of school mathematics arithmetic, algebra, and geometry. Might it be better for mathematics, and for students, if calculus were delayed until fewer (and more mature) students were enrolled ? Is calculus the right course to introduce the rigor of analysis ? Calculus not only represents the crowning achievement of the age of science, but it also serves as a phase transition in students mathematical passage from algebra to analysis. It is where the clear crystals of arithmetic, algebra, and geometry liquefy into the uid ideas of limits, derivatives, and real numbers. Earlier, when calculus served a very limited population of motivated and screened students who were preparing for careers in mathematically based elds, it functioned with three main goals : to teach students the mechanics of calculus, to introduce some of its myriad applications, and to introduce epsilon-delta arguments. But as calculus has come to serve a much more diverse clientele, this latter goal has gradually disappeared : syllabi now focus on tools and applications, assessments stress procedures. Thus most students who study calculus experience only results and applications, not methodology or foundations. Determining the appropriate balance between the pragmatic and intuitive on the one hand and the formal and rigorous on the other hand is largely a matter of values, a reection of the goals one wants to achieve in the course. Should mathematics stake its future on calculus ? Mathematics continues to hold a place of privilege in an increasingly crowded curriculum. Unlike most subjects, mathematics is universally compulsory for school children and widely recommended long after it becomes optional. Increasingly, its special place is being challenged, especially by information technology. Historically, mathematicians (and others) could easily make a strong case for retaining a focus on calculus not just for instrumental reasons of convenience but also on intellectual, cultural, pragmatic, social, and scientic grounds. But now calculus faces two signicant challenges from technology that can best even experts on most calculus tasks, and from signicant shifts in the balance of power among major elds of mathematics. Might there now be better exemplars to make the case for mathematics to an increasingly skeptical public ? Should calculus look the same everywhere ? How much local context should be reected in the content, applications, context, and pedagogy of a calculus course ? Can one expect the same textbook or syllabus to work as well in South Dakota as in South Africa ? This inquiry touches on one of the deepest epistemological issues of mathematics education the degree to which mathematics should be taught as a pristine objective discipline independent
207
of local culture and customs, or as one of many subjective aspects of culture that is tightly bound to problems of local interest. The former emphasis tends in the direction of pure mathematics, the latter in the direction of quantitative literacy. At risk of oversimplication, one might say that the former reects an elitist perspective, the latter a populist agenda. Which is better for the future of mathematics ? How much technology should be taught in calculus class ? As bank clerks use calculators (rather than traditional algebra) to compute interest and loan payments, so engineers use professional computer software (rather than traditional calculus) to solve analysis problems. These tools include symbol manipulating software, interactive visualization, mathematically active notebooks, applets, and modern web-based communication tools. In this age, calculus students can rightly expect not only to learn traditional procedures, concepts, and applications, but also to acquire modest uency in using the professional tools by which calculus is now practiced. Notwithstanding these legitimate expectations, there remains a fundamental question of purpose and value that needs to be thoughtfully addressed : Should calculus in school prepare people technologically for calculus at work, or should it instead focus primarily on understanding fundamentals as preparation for further study ? One cannot rightly assess the success of a course or program, much less design appropriate strategies, without rst reaching agreement on its goals. Should computer notation become part of standard calculus instruction ? As silicon computers began taking on the task formerly carried out by human computers, programmers ran right into one of the major roadblocks that has historically created so much difculty for mathematics students : ambiguous and two-dimensional notation. To make things work, programmers introduced new notation that could be typed on a standard keyboard and interpreted unambiguously by software that lacked human intelligence for guessing meaning from context. This new notation is ubiquitous : for multiplication, for exponents, : for ranges of summation, : for assignment, etc. Yet mathematics texts continue, for the most part, to use only traditional notation that evolved to meet nineteenth century needs when computers were nowhere in sight. Should mathematics adjust its standards so that students learn a single unambiguous system that can be communicated readily by e-mail ? Might the clarity of notation that helped computers also help students ? What is the appropriate balance of content and context ? Discussions of mathematics curricula often take place along two relatively independent dimensions : more or less (abstract) mathematical content, and more or less

208
L. A. STEEN
(authentic) worldly context. Different approaches to calculus can readily be plotted on a content-context plane some are high in one dimension and low in another, some are low in both (because they focus mostly on mechanics at the expense of both content and context). Few, perhaps none, are high in both content and context because that would require far more time and effort than students ordinarily have available. Choices must be made to achieve a suitable balance. Like so many other questions about calculus, the answer to this question reveals more about the values of the person who provides the answer than it does about the nature of the subject. Should theories of cognition inuence the way analysis is taught ? For at least a decade, if not longer, those who actively pursue research in mathematics education have been exploring cognitive issues related to learning calculus. Much attention has been given, for example, to the APOS stages in students developing grasp of the concept of function : these are rst seen as actions (calculating values), then as processes (reecting on actions collectively), then as objects (encapsulated processes), and nally as schemata (classes of objects dened by shared properties). Functions have also been studied as representations of correspondences and of covariation among related variables, ideas that continually evolve in students minds [Harel & Dubinsky 1992]. Not only functions but also other familiar mathematical objects the number line, the meaning of equality, the idea of area, even the idea of number itself undergo successive reconstruction as students pass through the phase transition between algebra and calculus [Tall 1991]. The issue I suggest for investigation is not which of these theories can be conrmed, but whether and to what degree any of these theories have signicant impact on the effectiveness of teaching. Should it matter what theory of cognition, if any, a calculus teacher believes ? Should mathematicians pay attention to what neuroscientists are learning about the mental constructs of mathematics ? Heretofore, all theories of mathematical cognition have been based exclusively on behavioral evidence rather like theories of illness before the germ theory of disease. Now, however, physical and biological evidence from neuroscience is beginning to appear, evidence about electrical activity in the brain when it engages in mathematical thought [Butterworth 1999 ; Dehaene 1997 ; Lakoff & N unez 2000]. Although this evidence is still primitive, it may help suggest or clarify possible explanations for what we observe about the struggles students have in learning mathematics. For example, it now seems clear that the well-known difculty of mastering the multiplication table has physiological roots since the words for numbers are handled in a different part of the brain than
209
other words. Might we someday learn something about how the brain deals with limits or other abstractions of analysis ? Might such insight improve the teaching and learning of calculus ?
7. SUMMING
UP
The challenges facing analysis education at the beginning of the twentyrst century, and the second century of LEnseignement Math matique, are e quite different from those one might have listed fty or one hundred years ago. Calculus is now relatively more important in more students lives even while it represents a relatively smaller fraction of the expanding mathematical universe. Calculus is no longer, as it once was, the unchallenged tool of applied mathematics ; statistics, linear algebra, and combinatorial methods all have strong claims to the crown of mathematical utility. Moreover, calculus is not even the only option for accomplishing its etymological purpose, namely calculating with what used to be quaintly called innitesimals. Computers, thank you, can now do all those things perfectly well. As we enter the new century, I suggest that we must be forthright in challenging all historic assumptions about the role of calculus in the mathematics curriculum. Take nothing for granted ; put all options on the table. If calculus were not now entrenched, would we choose to give it such high priority in such a crucial place in the curriculum ? Are there better choices to help attract students to the advanced study of mathematics ? What benets might ensue if secondary school mathematics were freed of the burden of preparing all students for calculus ? Answers we give to these questions may well determine the fate of higher mathematics education, whether it will be seen as the relic of a glorious past that has now been overtaken by new subjects such as information technology, or whether it will be seen, as it has been for most of the last century, as a conveyor of skills and understandings that are crucially important for all educated people. We dare not take the future for granted.
REFERENCES ADELMAN, C. A parallel universe : certication in the information technology guild. Change 32 : 3 (May/June 2000), 2029. ASKEY, R. A. The third mathematics education revolution. In : Contemporary Issues in Mathematics Education, ed. E. A. Gavosto, St. G. Krantz and W. McCallum, 95107. Cambridge University Press, New York, 1999.
210
L. A. STEEN
BOYER, E. L. Scholarship Reconsidered : Priorities of the Professoriate. The Carnegie Foundation for the Advancement of Teaching, Princeton, 1990. BUTTERWORTH, B. What Counts : How Every Brain Is Hardwired for Math. Free Press, New York, 1999. DEHAENE, S. The Number Sense : How the Mind Creates Mathematics. Oxford University Press, New York, 1997. DOUGLAS, R. G. (ed.) Toward a Lean and Lively Calculus. Mathematical Association of America, Washington, 1986. DUBINSKY, E., J. KAPUT and A. SCHOENFELD (eds). Research in Collegiate Mathematics Education (Vols IIII). (Conference Board of the Mathematical Sciences : Issues in Mathematics Education.) Amer. Math. Soc., Providence, 1994, 1996, 1998. GOLD, B., S. J. KEITH and W. A. MARION (eds). Assessment Practices in Undergraduate Mathematics. Mathematical Association of America, Washington, 1999. HAREL, G. and E. DUBINSKY (eds). The Concept of Function : Aspects of Epistemology and Pedagogy. Mathematical Association of America, Washington, 1992. HODGSON, B. R. Pr face. In : Calcul diff rentiel (Le Projet Harvard), ed. D. Hughese e Hallett, A. M. Gleason et al., iiiiv. Cheneli` re/McGraw-Hill, Montr al, 2000. e e JOINT POLICY BOARD FOR MATHEMATICS. Recognition and Rewards in the Mathematical Sciences. Amer. Math. Soc., Providence, 1994. KAPUT, J. J. and E. DUBINSKY (eds). Research Issues in Undergraduate Mathematics Learning. Mathematical Association of America, Washington, 1994. LAKOFF, G. and R. NUNEZ. Where Mathematics Comes From : How the Embodied Mind Brings Mathematics into Being. Basic Books, New York, 2000. LOFTSGAARDEN, D. O., D. C. RUNG and A. E. WATKINS. Statistical Abstract of Undergraduate Programs in the Mathematical Sciences in the United States : Fall 1995 CBMS Survey. Mathematical Association of America, Washington, 1997. NISS, M. Key Issues and Trends in Research on Mathematics Education. In : Proceedings, 9th International Congress on Mathematics Education (forthcoming). ROBERTS, A. W. (ed.) Calculus : The Dynamics of Change. Mathematical Association of America, Washington, 1996. SCHOENFELD, A. (ed.) Student Assessment in Calculus. Mathematical Association of America, Washington, 1997. SPENCE, L. H. Architect of AP Action Plan outlines challenges. The College Board News 28 : 3 (June 2000), 19. STEEN, L. A. (ed.) Calculus for a New Century : A Pump, Not a Filter. Mathematical Association of America, Washington, 1988. STEWART, D. M. (ed.) Advanced Placement Yearbook. The College Board, New York, 1997. TALL, D. (ed.) Advanced Mathematical Thinking. Kluwer, Dordrecht, 1991. TUCKER, A. C. and J. R. C. LEITZEL (eds). Assessing Calculus Reform Efforts. Mathematical Association of America, Washington, 1995. WU, H-H. The mathematics education reform : Why you should be concerned and what you can do. Amer. Math. Monthly 104 (1997), 946954.
REACTION Enseignement et apprentissage de lanalyse : que retenir du pass pour penser le futur ? e par Mich` le ARTIGUE e
En r action aux trois contributions sur lenseignement et lapprentissage de lanalyse e au 20e si` cle, pr sent es respectivement par Jean-Pierre Kahane, Man-Keung Siu et e e e e ` Lynn Arthur Steen, jai et particuli` rement sensible a plusieurs ph nom` nes. e e e e Jai et dabord frapp e par le fort contraste existant entre la situation au d but du e e 20e si` cle et la situation actuelle. Le d but du si` cle voit lentr e modeste mais efcace e e e e de lenseignement du calcul diff rentiel et int gral dans les lyc es, dans de nombreux e e e pays, lenthousiasme des pionniers, le sentiment quils expriment avec force davoir r ussi dans leur entreprise. La situation actuelle semble au contraire caract ris e par e e e un sentiment croissant de crise que les diff rentes r formes entreprises dans le monde e e ` ` e entier ne parviennent ni a enrayer, ni m me a r duire. e Le second point qui ma frapp e, cest la permanence tout au long du si` cle des e e m mes questions, en particulier les suivantes : Comment r percuter dans le choix des e e contenus denseignement l volution scientique de ce domaine, celle des rapports e entre math matiques et autres disciplines ? Comment prendre en compte la diversit e e e des besoins des el` ves et etudiants ? Comment trouver un juste equilibre entre intuition et rigueur dans les premiers contacts avec le monde de lanalyse ? Mais la pr sentation de Lynn Steen a par ailleurs bien montr quau del` de e e a ces permanences, emergent des questions nouvelles, comme les suivantes : Comment penser lenseignement de lanalyse aujourdhui compte tenu de la massication de lenseignement secondaire, et maintenant universitaire, et de lh t rog n it croissante ee e e e des etudiants qui en r sulte ? Comment prendre en compte dans cet enseignement e l volution technologique ? Quoffre, pour penser lenseignement et lapprentissage de e lanalyse, la recherche didactique qui sest d velopp e maintenant depuis plus de vingt e e ans dans ce domaine et que pourrait-on en attendre de plus ?
212
M. ARTIGUE
Ces questions nous montrent clairement que les contextes changent et que les solutions aux probl` mes denseignement et dapprentissage de lanalyse qui se posent e e aujourdhui ne peuvent etre simplement emprunt es au pass , quelles quaient et ses e e r ussites. e ` Dans ce texte, je souhaite contribuer a la discussion sur ces questions, en respectant e lapproche historique qui a et celle du Symposium et me semble etre particuli` rement e productive. Quand on r chit sur les questions denseignement et dapprentissage, se e e ` pencher sur lhistoire est quelque chose dessentiel. Cela aide a percevoir lorganisation actuelle dun enseignement, ses choix et ses valeurs, comme le r sultat dune longue e histoire partiellement d pendante des cultures math matiques et surtout educatives dans e e ` lesquelles cette histoire sest inscrite. Cela aide a mieux comprendre la nature et la force des contraintes avec lesquelles les syst` mes educatifs doivent compter, les r sistances e e ` qui souvent sopposent aux changements curriculaires, les effets a moyen ou long terme ` ` souvent inattendus de telle ou telle d cision. Cela aide enn a int grer a la r exion e e e la vision dynamique n cessaire pour penser lavenir. e Je reviendrai donc dans un premier temps sur le contraste soulign pr c demment et e e e essaierai de tirer quelques lecons de son analyse. Dans un deuxi` me temps, jessaierai e ` de contribuer a la r exion sur les questions que pose aujourdhui lenseignement des e d buts de lanalyse au lyc e, en identiant notamment les ressources offertes par les e e recherches didactiques men es dans ce domaine. e
REACTION Learning and teaching analysis : What can we learn from the past in order to think about the future ? by Mich` le ARTIGUE e
AN
EVIDENT HISTORICAL CONTRAST
What we observe at the beginning of the 20th century is the timid introduction of differential and integral calculus in curricula which are still dominated by geometry. For instance, in the new 1902 French syllabus for scientic sections at grade 12, analysis which is included in the algebra part takes up only half a page from a total of nine. The teaching of analysis had limited ambitions : developing an efcient elementary knowledge of calculus. These ambitions met both rigour and usefulness requirements as they were expressed at that time : this calculus was free from the metaphysics of innitesimals, it offered unied methods for computing geometric, mechanical, and physical magnitudes, for studying variation and motion. Teachers were provided with efcient mathematics resources which introduced them to the new contents to be taught, beyond the sole student-textbooks. We must add that high-school students were part of the social elite and constituted a rather homogeneous population : for instance, in France less than 10% of the population entered the high-school system at that time [Belhoste 1996]. With the turn of the mid-century came a substantial evolution. Man-Keung Siu evoked his personal experience as a secondary-school student in the early sixties, when calculus teaching was a stabilized corpus, taught in an intuitive and informal way which emphasized calculation rather than conceptual understanding. As he said, he enjoyed doing calculus problems, applying the
214
M. ARTIGUE
taught recipes for calculating the escape speed of a rocket for instance , but he did not really understand what was going on. With the new maths movement, new ambitions developed for the teaching of mathematics and, as a specic case, for the teaching of analysis. Beyond efcient calculus, what was now being pursued, in many countries including France, was an approach to analysis as a structured theoretical eld where concepts were dened formally and where approximation played a fundamental role [Artigue 1996]. Analysis became independent of algebra and the balance between analysis and geometry progressively changed in the high-school curriculum. The rejection of the structural values of the new maths reform, when it occurred (as it did in France in the early eighties), did not alter this tendency. Reducing the role given to algebraic structures and to the settheoretic dimensions which had been emblematic of that reform, cleared some curricular space which analysis tended to occupy. It is important to remark that this increasing role of analysis took place in a mathematical environment which was becoming less extensive (for instance, in France, astronomy, mechanics, and kinematics had progressively disappeared from the high-school curriculum) and where the connections with other scientic disciplines, which were deemed so important at the beginning of the century, were gradually fading out. The present situation, as stressed by L. Steen, is characterized both by an over-representation of analysis in many high-school curricula and by a general feeling of crisis which seems to transcend cultural differences. In his text L. Steen has described the US situation : the teaching of analysis is optional at the secondary level ; what is taught is still mainly calculus, but the teaching content is ever more challenged by the advances of computer technology, which seems able to take charge of the full range of abilities and competence expected from the students. In France, the context is not quite the same. Analysis teaching is compulsory, its aims go beyond mere calculus, even if theorization and formalization have been progressively reduced during the last twenty years. In fact, our educational system, facing the constraints of mass teaching, is much less able to foster the epistemological values it claims to develop. These difculties generate strong frustrations. Moreover, the predominant role of analysis in the syllabus, as is proved by its key position in the baccalaur at 1 ), is challenged by other mathematical topics : e mainly, statistics and probability theory or discrete mathematics. Statistics and probabilities are more and more considered to be necessary parts of the mathematical culture of educated citizens living in a democratic country
1
) the national examination ending high-school studies, which allows entrance into university.
ANALYSIS
215
like France. Discrete mathematics is strongly favoured by the technological evolution and the connections between mathematics and computing. Thus, the crisis does not look everywhere the same, but it is a perception which transcends cultural differences. There is an obvious necessity of rethinking the teaching of analysis at the secondary level. How relevant will it be at the turn of the new century if we take into account the current institutional constraints (especially mass education and the resulting students heterogeneity), technological advances or the evolution of social and scientic needs ? These questions have been extensively addressed by L. Steen in his contribution. He emphasized that there is not a unique solution. In fact, each of the answers we can imagine necessarily includes some ideological choices which we have to integrate. Even though, as pointed out by M.-K. Siu, the fundamental difculties of the learning of analysis have not substantially changed, what has been stressed above shows also that radical changes in contexts and values prevent us from simply borrowing solutions from the past. SO,
WHAT CAN WE LEARN FROM THE PAST ?
What was perceived as an adequate solution at the beginning of the 20 th century or in the eighties cannot be seen as a possible solution now. But the past certainly helps to question the present tendencies, values, and choices. We will use history within this perspective in what follows, when evoking three separate points : the respective role of geometry and graphical representations in the meaning of analytic concepts, the transition between algebra and analysis, and technological issues. HOW DO THE CONCEPTS Up to the seventies, as was stressed by J.-P. Kahane and M.-K. Siu, the geometrical and kinematic culture of students certainly played an essential role by allowing students to give some meaning to computational practices which were far from being clearly understandable. Today after the theoretical attempts of the new maths reform and the reduction of the mathematical students landscape, graphical visualizations offered by calculators and computer software are apparently supposed to play a similar role. All over the world, the curricula emphasize the necessity (in the rst contacts with functions and analysis) of interconnecting the algebraic approaches, which have been predominant for a long time, with numeric and graphical ones [Robert & Speer 2001]. The triad numericgraphicalgebraic has become emblematic of different
OF ANALYSIS BECOME MEANINGFUL FOR STUDENTS ?
FROM
GEOMETRY TO GRAPHIC VISUALIZATIONS :
216
M. ARTIGUE
projects such as the Harvard Calculus Project in the USA. Didactic research, more globally, tends to offer arguments in support of such a strategy, by showing the role played in conceptualization processes by a certain amount of exibility between different semiotic registers of representation, as is for instance extensively discussed in [Duval 1995]. Beyond that, the dynamic character of the graphical visualizations produced by computer technology, and the possibility of connecting such visualizations with mechanical devices in order to simulate motion and also to analyse and predict the effects of changes in speed and acceleration, offer us, today, new ways of exploiting the potential provided in the past by classical kinematic perspectives. Research conducted by J. Kaput and his colleagues [Kaput 1992], in connection with the software MathCars, is a good example : this pioneering work showed how a rst qualitative contact with variational issues and concepts can be introduced very early, at a time when students do not yet master the algebraic techniques that are generally considered a prerequisite to learning in this area. Up to what point are graphic visualizations efcient tools for supporting the development of mathematical knowledge in analysis ? What kind of tasks, of problematic situations can allow the students to maximize the expected benet of these techniques ? What are their real potential and limits ? Are they adequately exploited in current teaching with graphic calculators ? Even if research allows us to approach these questions better today, they remain widely open. Moreover, there is no clear evidence that graphic calculators, as they are normally used in high-school educational contexts, offer the possibility of a deep conceptual understanding in analysis. Graphical visualization is not meaningful per se. Most research results show that the students spontaneous interpretation of graphical phenomena may be quite different from what is expected by the teacher, who reads these phenomena through his own mathematical background (see for instance [Trouche 1994]). Loading graphical representations with mathematical substance requires specic apprenticeship which is often underestimated. It is far from easy to connect graphical explorations and symbolic representations in order to go beyond the evident limitations of graphical descriptions (for instance, they do not allow the control of approximation orders, which is essential in analysis). Research certainly provides some useful resources for this work (see [Tall 1996] for a synthetic view), but, as is often the case in education, what is discovered in an experimental setup does not seem easy to share and generalize. The fact that most calculus reform projects in the US are not closely linked to research, as evidenced by N. Speer in a recent survey [Robert & Speer 2001], is not anecdotal from this point of view. We will come back to this point further on.
ANALYSIS
217
THE TRANSITION BETWEEN ALGEBRA AND ANALYSIS. In his contribution L. Steen has evoked this transition and stressed its logical demands. In this paragraph, I would like to evoke briey the diversity of the reconstructions at play in this transition and the long-term evolution they require. I will then discuss the curricular implications that such considerations can have. Entering a new eld of mathematical knowledge generally requires some cognitive reconstructions that are not easy to achieve. This cognitive cost is very often underestimated by educational systems ; generally, most of these reconstructions are left to the personal responsibility of students and this is the origin of persistent difculties and failures. Such difculties have been extensively investigated in the case of the transition between arithmetic and algebra [Bednarz, Kieran & Lee 1996]. Surely, analysis also obeys the general situation. Not all the cognitive reconstructions that are required for entering the eld of analysis are of the same nature [Artigue 2001a]. Some of them deal with mathematical objects or practices which are familiar to students before they get acquainted with the ofcial teaching of analysis. Some others are more internal to the eld of analysis. They can result from the polysemy of mathematical concepts (like the integral or the derivative) and from the fact that the different facets of such notions cannot all be introduced at the same time. Others still are linked to the fact that, as H. Poincar underlined at the e beginning of the century in his famous lecture on mathematical denitions [Poincar 1904], a mathematical concept cannot, as a rule, be introduced to e students from the beginning in its most achieved form. Different levels of conceptualization have to be progressively reached, each one corresponding to a kind of balance which is partially broken when a more complex level is striven for. In order to analyse the transition between algebra and analysis, we shall need to go into these different categories, as the relationships between algebra and analysis depend on the levels of conceptualization that one adopts. For instance, when the teaching of analysis begins at high-school level, students have already met irrational numbers and functions. These objects are implied in their mathematical practices, and the conceptions they generate emerge from these practices. Mathematical practices can be described, as Chevallard does in his anthropological approach to mathematics education [Chevallard 1992 ; Bosch & Chevallard 1999], in terms of praxeologies structured around tasks, techniques developed for solving these tasks, and the technological and theoretical discourses which are used to motivate, explain, and justify these techniques. Thus the real numbers that emerge from students practices are algebraic objects with a dense order, a geometric representation
218
M. ARTIGUE
(the real line), and numerical representations or approximations, among which the nite decimal expansions given by calculators play an essential role. As we said before, entering the eld of analysis requires some complex reconstructions. Research data show that, even if students declare that the real order is a dense order, they are tempted to think that some numbers can, in some sense, be successors. For instance 0.999 is very often considered to be the last number before 1. In fact, the symbolic notation evokes an innite process whose terms are all less than 1, and it looks as though students could not detach the result of the process from the process itself. Tests at the entrance of French universities have regularly shown that the majority of science students answered the question : what can you say about a and b if, for every positive integer n , a b 1 n ?, by saying that a and b are innitely close or successors. Such conceptions obviously make it difcult to produce or even understand proofs in analysis. Research also shows that the conceptions the students have of the real line are at variance with what is expected in analysis, at a certain level at least. A spontaneous conception of continuity develops on the basis of our physical experience with space and motion [Lakoff & Nunez 2000 ; Nu ez, Edwards & Matos n 1999]. The real line is thus at rst linked to the idea of continuous motion, of trajectory ; real points are not constitutive of it, they appear more as points on a pre-existing line, as milestones along a highway. Understanding the essential role played by completeness, perceiving continuous functions as functions that preserve the continuity of subsets, requires subsequent changes. When they begin to be taught analysis, students have already developed some experience with functions. Once more, their relations to functional objects emerge from the practices these objects have been involved in. Today, these practices generally favour an introduction to functions through modelling situations, combining numerical, graphical and algebraic work. Variational issues are also considered mainly by relying on numerical and graphical evidence, or on the properties of prototypical functions. These practices are of a point-wise or of a global nature. Entering the eld of analysis forces one to reconsider them by taking into account the localization of the perspective that is proper to analysis and the dialectic interplay that can be developed between local and global perspectives. Linearity, which was a global concept, has for instance also to be thought of as a local property shared by a large class of functions, and students need also to understand how such local conditions can fully characterize mathematical objects, as is the case with differential equations.
ANALYSIS
219

Even without entering formal analysis and - proofs, algebraic practices have to be reconsidered. In the previous algebraic work, all the different components of an algebraic expression were given the same weight ; solving an equation meant nding all the numbers satisfying that equation. In analysis, the management of algebraic expressions has to take into account the different orders of magnitude of the terms and to look for what is predominant and what can be neglected. Working with inequations is no longer playing the same game ; it means combining this differentiated treatment of expressions with the play on intervals or neighbourhoods induced by the local perspective. The technical work thus deeply changes and becomes more complex, as well as the heuristics and the control processes. The transition towards higher levels of conceptualization and what is generally called formal analysis where objects are dealt with on the basis of their formal denitions needs reconstructions of a different nature. As stressed by many researchers [Tall 1996 ; Robert 1998] and evidenced by empirical results, there is a deep gap between what I have evoked above and formal analysis, even if what I have evoked goes beyond mere calculus in the classical sense. Formalized concepts, such as the formalized concept of limit, cannot be built in continuity with the intuitive sources linked to social and physical experience. In essence they are proof-generated concepts [Lakatos 1976] answering foundational needs. As pointed out by Robert [1998], understanding such mathematical needs is far removed from the mathematical culture of secondary and even university students, and some specic strategies have to be designed in order to allow them to penetrate this new culture. Taking all these facts into consideration obliges us to think in new terms the problem of the secondary teaching of analysis. Even if we are convinced that, for cultural reasons, students cannot leave secondary school without developing some contact with the ways innity and innite processes are approached by mathematicians, analysis is not necessarily the most appropriate domain for all categories of students. Entering the epistemological eld of analysis as we perceive it today, whatever choice we make, is costly and involves longterm processes. This eld can certainly be approached in different ways, as we have tried to show. High-school teaching, without reducing to mere algebraic calculus, nor looking for inaccessible aims such as formal levels of conceptualization, can offer a coherent way that initiates the necessary reconstructions. Nevertheless, the cost of such an initiation should not be minimized and, for students who do not require such expertise for their professional future, the compulsory character of analysis at high-school level should certainly not be taken for granted.
220
M. ARTIGUE
TECHNOLOGICAL ISSUES. As regards this essential point, I wish to emphasize that we still live under a vision which, in my opinion, prevents us from addressing technological issues efciently. The values of mathematics teaching and learning are today dened essentially by referring to paper and pencil learning environments and practices, as though mathematical values could ideally be thought of in an absolute way, without taking into account the current instruments of mathematical work, the way these shape our mathematical practices and change our mathematical needs. Technology is expected to improve learning processes, students motivation, etc., and its educational legitimacy seems to depend on its ability to full these requirements. Research work in analysis has considered technology within different complementary perspectives : through a programming perspective, programming activities being seen as a way to solve the process/object gap (the development and use of the language ISETL in the framework of APOS theory [Dubinsky & Mac Donald 2001] can be considered a paradigmatic example of this perspective) ; through a semiotic perspective, using technology as a means of visualizing mathematical objects and connecting semiotic representations, as is the case for instance in [Borba & Confrey 1996] ; through an experimental perspective, technology being seen as a means of developing more experimental approaches involving motion, simulation, and problems coming from different scientic contexts. This has produced interesting results and engineering designs but, on the whole, it did not seriously challenge the traditional vision, which is therefore poorly sensitive to crucial issues such as : the computer transposition of mathematical knowledge [Balacheff 1994] and the ways in which such a transposition can affect our relationship to mathematical objects and our mathematical needs ; the complexity of the instrumental genesis through which an artefact such as a calculator or a piece of software becomes an efcient mathematical tool with its specic mathematical needs. As evidenced by recent research on computer algebra systems (CAS) [Guin & Trouche 1999 ; Lagrange 1999 ; Defouad 2000], it seems especially important to develop this kind of awareness in a domain such as analysis. In fact, CAS were not designed for educational purposes but for professional use ; however, they are likely, in the near future, to become tools as ordinary as graphic calculators are today in secondary schools.
ANALYSIS
221
Thus, as was stressed by Lynn Steen, conceiving the future of analysis in secondary schools today requires an important change of perspective :
Mathematics educators now must deal not only with the question of how technology can enhance (or impede) mathematical learning and how technology changes priorities for mathematics content, but also how technology changes the way mathematics is expected to be performed. Technologys challenge to mathematics education is now about ends as well as means. [Steen 2003, 194]
CONCLUDING
REMARKS
As I have tried to demonstrate, educational research has a lot to offer to our reection through the results it has obtained : about students conceptions and difculties, with evident coherence in the results all over the world, in spite of natural differences in cultural and institutional contexts, about the dysfunctionings of educational systems in that area, the nature and strength of the constraints that oppose to evolution, and about ecological and viability issues, but also through : the theoretical frames and concepts it has developed for thinking about these questions, such as the APOS theory quoted by L. Steen, which put to the fore the qualitative gap linked to the transition between the process and object conceptions of mathematical objects, which plays an essential role in learning processes in analysis, the engineering designs it has developed and tested, such as for instance the design elaborated by the group AHA (Heuristic Approach to Analysis) in Louvain-la-Neuve [Groupe A.H.A. 1999], which offers a rather ambitious programme for high-school level relying on a strong epistemological analysis and trying to progress from mental objects, with the meaning given to this term by H. Freudenthal, to constructed objects. But research has also evident limits and certainly does not allow one today to answer many of the issues raised by L. Steen. Some of its limitations can be regarded as contingent, like those related to the poor development of research in specic areas such as the teaching of analysis in service courses. Others are not so contingent but are linked to the complexity of educational processes. Research can inform reection and choices, it cannot guide them in a mechanical way, neither can it be asked to take in charge, in a scientic way, questions which mainly depend on ideological or political values. Moreover, its results, contrary to mathematical theorems, are time and space dependent
222
M. ARTIGUE
and, even if coherent, they are not established within one unied paradigm and do not lead to identical didactic choices [Artigue 2001b]. These characteristics certainly make their transmission and use difcult, beyond local communities of research, and contribute to explaining the limited impact they have had up to now on innovation and reform projects.
REFERENCES ARTIGUE, M. R formes et contre-r formes dans lenseignement de lanalyse au lyc e. e e e In : B. Belhoste et al. (eds), Les sciences au lyc e un si` cle de r formes des e e e ` e math matiques et de la physique en France et a l tranger, 197217. Vuibert, e Paris, 1996. L volution des probl matiques en didactique de lanalyse. Recherches en Didace e tique des Math matiques 18.2 (1998), 231262. e [2001a] What can we learn from educational research at the university level. In : D. Holton et al. (eds), The Teaching and Learning of Mathematics at University Level : an ICMI Study, 207220. Kluwer, Dordrecht, 2001. [2001b] Lentr e dans le champ conceptuel de lanalyse : r formes curriculaires, e e recherches didactiques, o` en est-on ? In : T. Assude et B. Grugeon (eds), Actes u du s minaire national de didactique des math matiques, ann e 2000, 277301. e e e IREM Paris 7, Paris, 2001. BALACHEFF, N. Didactique et intelligence articielle. Recherches en Didactique des Math matiques 14.1 (1994), 942. e BEDNARZ, N., C. KIERAN and L. LEE (eds). Approaches to Algebra : Perspectives for Research and Teaching. Kluwer, Dordrecht, 1996. BELHOSTE, B. R former ou conserver ? La place des sciences dans les transformations e de lenseignement secondaire en France (19001970). In B. Belhoste et al. (eds), Les sciences au lyc e un si` cle de r formes des math matiques et de la physique e e e e ` e en France et a l tranger, 2738. Vuibert, Paris, 1996. BORBA, M. and J. CONFREY. A students construction of transformation of functions in a multiple representational environment. Educational Studies in Mathematics 31.3 (1996), 235268. BOSCH, M. et Y. CHEVALLARD. La sensibilit de lactivit math matique aux ostensifs. e e e Objet d tude et probl matique. Recherches en Didactique des Math matiques e e e 19.1 (1999), 77124. CHEVALLARD, Y. Concepts fondamentaux de la didactique : perspectives apport es par e une approche anthropologique. Recherches en Didactique des Math matiques 12.1 e (1992), 77111. DEFOUAD, B. Etude de gen` ses instrumentales li es a lutilisation dune calculatrice e e ` symbolique en classe de premi` re S. Th` se de doctorat. Universit Paris 7, 2000. e e e DUBINSKY, E. and M. MC DONALD. APOS : a constructivist theory of learning in undergraduate mathematics education research. In : D. Holton et al. (eds), The Teaching and Learning of Mathematics at University Level : an ICMI Study, 275282. Kluwer, Dordrecht, 2001. DUVAL, R. Semiosis et pens e humaine. Peter Lang, Berne, 1995. e
ANALYSIS
223
` GROUPE A.H.A. (eds). Vers linni pas a pas. Approche heuristique de lanalyse. De Boeck, Bruxelles, 1999. GUIN, D. and L. TROUCHE. The complex process of converting tools into mathematical instruments : the case of calculators. The International Journal of Computers for Mathematical Learning 3 (3) (1999), 195227. KAPUT, J. Technology and mathematics education. In : D. Grouws (ed.), Handbook of Research in Mathematics Teaching and Learning, 515556. Macmillan, New York, 1992. LAGRANGE, J. B. Techniques and concepts in pre-calculus using CAS : a two-year classroom experiment with the TI92. The International Journal for Computer Algebra in Mathematics Education 6.2 (1999), 143165. LAKATOS, I. Proofs and Refutations, the Logic of Mathematical Discovery. Cambridge University Press, 1976. LAKOFF, G. and R. NUNEZ. Where Mathematics Comes From : How the Embodied Mind Brings Mathematics Into Being. Basic Books, New York, 2000. NUNEZ, R., L. D. EDWARDS and J. P. MATOS. Embodied cognition as grounding for situatedness and context in mathematics education. Educational Studies in Mathematics 39 (1999), 4565. POINCARE, H. Les d nitions g n rales en math matiques. LEnseign. Math. 6 (1904), e e e e 257283. ` ` ROBERT, A. Outils danalyse des contenus a enseigner a luniversit . Recherches en e Didactique des Math matiques 18.2 (1998), 139190. e ROBERT, A. and N. SPEER. Research on the teaching and learning of calculus / elementary analysis. In : D. Holton et al. (eds), The Teaching and Learning of Mathematics at University Level : an ICMI Study, 283299. Kluwer, Dordrecht, 2001. STEEN, L. A. [2003] Analysis 2000 : challenges and opportunities. This volume, 191 210. TALL, D. Functions and calculus. In : A. J. Bishop et al. (eds), International Handbook of Research in Mathematics Education, 289325. Kluwer, Dordrecht, 1996. TROUCHE, L. Calculatrices graphiques : la grande illusion. Rep` res IREM 14 (1994), e 3955. Enseigner en terminale S avec des calculatrices graphiques et formelles. IREM de Montpellier, 1998.
GENERAL DISCUSSION (reported by Marta MENGHINI)
Following Mich` le Artigues reaction, there was a lively debate about e analysis and the problems facing the teacher and the learner. The two different aspects of analysis, namely its usefulness in applications and as a service subject, and its conceptual role, were illuminated by the discussion about how, and to whom, the subject should be taught. The tension between analysis as a tool and analysis as a branch of pure mathematics was illustrated in an extended contribution by Jean-Pierre Kahane, devoted to the historical development of Fourier analysis. Fourier rst adapted trigonometric series to study the propagation of heat. Analysis was applied to physical investigations in order to nd general procedures aimed at calculating the sums of series. The usefulness of series is not their convergence, but their power to compute. During the 19th century Fourier series became something else : they provided the opportunity for Dirichlet to state the rst important theorem about the convergence of functions and also for Riemann to produce his idea of integrals. Finally, in the spirit of the end of the 19th century, what was taught as Analysis became Foundations of Analysis. Later Bourbaki considered fundamental structures to be paramount in mathematics, and so analysis was then taught as General Topology. And now ? We are turning again to discrete Fourier transforms, now adapted to new tools and applied to many new elds. The major part of the discussion revolved around the teaching of analysis and the difculties this entails, which are necessarily linked to the conceptual problems already indicated. Certainly analysis is a mathematical tool which is useful to non-mathematicians. Hence it is important to note that if mathematicians offer only traditional (pure) analysis then the subject will be taught by scientists and engineers to their students in the way that suits them best. Furthermore, motivation for learning the subject can be found in its possible applications and this should be used by mathematicians.
ANALYSIS
225
Most of the participants who spoke in favour of a more applied view of analysis were referring to university level courses, but one speaker made the point that, at school level too, it is exactly the awareness of the applications of analysis, particularly its physical interpretations (which are too often absent), that helps to clarify the general meaning of analysis. And Ubiratan DAmbrosio suggested that the idea of mathematics in a lab should now be renewed, with the addition of computer facilities, as a step towards more theoretical reections. New technology offers new opportunities in teaching and learning analysis and many participants were keen to see the development of more conceptual and structured tasks, in order to develop thinking in analysis ; the issues regarding conceptual and procedural knowledge should not be neglected, and this is particularly important now, in the era of CAS (Computer Algebra Systems) and DGS (Dynamic Geometry Software), when using technology and computer transposition. Even the practical problem of determining areas takes on a conceptual aspect if the historical approach of teaching integral calculus before derivatives is followed. Trevor Fletcher maintained that integration seems to be easier than differentiation, because people studied it successfully 2000 years before. He remembered that some decades ago there were people who seriously considered developing elementary analysis by dealing with integration before differentiation, assuming that would be a better motivation for the learning of analysis. His question whether this is an issue today produced many reactions. Jean-Pierre Kahane recalled that at the beginning of the 20th century the general feeling among Italian mathematicians had been to begin with integration and there had been discussions about how this could be done. But he pointed out that for computating integrals one needs derivatives, regardless of the fact that integration was rst to appear in the work of Archimedes. A balance is therefore necessary. A possibility could be to introduce integration rst, presenting areas and volumes as motivation and then, soon after, to introduce derivatives in order to perform computations. Similarly, another speaker recalled that in the 60s Tom Apostol published textbooks for teaching calculus, starting with the integral and later making the connection with derivatives. Conceptually it was very nice, but there were problems with students studying physics, and the order was reversed. Many of these approaches, which sound attractive to mathematicians, seem to have been abandoned because of the difculty of not having developed computational tools early enough for the student, in other words because of the need to preserve analysis as a service subject.
226
GENERAL DISCUSSION
The question whether the derivative or the integral is the more fundamental concept of analysis was present in a number of contributions to the discussion. Certainly the idea of area seems more fundamental but this raises the conceptual problem of area itself. Given the concept of integration and the concept of real number we could revisit concepts which have been accepted before, such as the fact that the area of a rectangle is given by the product of its two sides. The question is whether this is a theorem to prove, or a well-stated denition. It is also important not to lose sight of the difculties inherent in the teaching of analysis. Geoffrey Howson recalled that analysis was never easy to teach even at university level. In the 60s those people who were developing the SMP (School Mathematics Project) were faced with the problem of how to approach the concept of differential, and what was the key idea they wanted to be left in the students mind. On this there was no agreement. There were two different approaches : one went through developing functions and mappings, the other aimed to emphasize the role of approximation. Unfortunately, whatever the approach, the concept of a limit was soon needed, after which the derivative was to be introduced, and nally algebraic techniques in analysis followed. Which approach to use still seems to be the crux of the problem, together with the problem of notation. Another participant reminded the audience that the fact that analysis is introduced so late in the curriculum seems to render unnecessary the use of a simpler language, which raises the question of the advisability of looking at it in a more elementary way. But the problem of what sort of analysis, which approach to use and, fundamentally, whether to begin with the derivative or the integral, still remains and is by no means new. Gert Schubring concluded the discussion by noting that the question of which parts of calculus should be introduced in schools, remains a problem today just as it had been at the beginning of Kleins reform movement. By common agreement, analysis should not be reduced to a set of formal techniques, so how can calculus be introduced in a way that provides conceptual understanding ? A nal remark, endorsed by the participants, was that the question of identifying conceptual tasks in analysis would be an appropriate topic for a special issue of LEnseignement Math matique. e
APPLICATIONS OF MATHEMATICS : MATHEMATICS AS A SERVICE SUBJECT
LES DEBATS AUTOUR DES APPLICATIONS DES MATHEMATIQUES DANS LES REFORMES DE LENSEIGNEMENT SECONDAIRE ` ` AU DEBUT DU VINGTIEME SIECLE The debates around mathematics applications in the reforms of secondary teaching at the beginning of the twentieth century by Philippe NABONNAND
One of the primary aims for establishing the International Commission on Mathematical Instruction (ICMI) was to publish contributions on the reforms of mathematical instruction. The journal LEnseignement Math matique soon became e ICMIs ofcial organ and a forum for debating questions pertaining to pedagogy, didactics, curriculum planning, and the organization of mathematical instruction. At the close of the 19th century, the advances of the sciences and the growth of industrial society showed how necessary it had become to reform mathematics education. The demand for well-trained technicians and engineers led to the introduction of technological and vocational education at the secondary level and above. The question as to the position of mathematical training within an increasingly practical educational system thus arose. Applied mathematics, or more precisely applications of mathematics, was the main focus for reformers. Most of the authors who published articles in LEnseignement Math matique were in favour of this spirit of reforms. e The chief objective of the present paper is to use the articles published in the periodical to provide an understanding of the position which applications of mathematics were to have in the new curricula emerging at the beginning of the 20 th century. Reformers argued in favour of introducing new topics of applied mathematics, advocating less theoretically oriented and more application-minded mathematical training. Surprisingly, there were no signicant developments in curricula along the lines of the intended changes, except for graphical and numerical methods. While reformers wanted mathematical training to meet the requirements of users and of the other sciences, it was, in their view, still far more important to introduce calculus into the curricula of secondary schools. Arguing in favour of introducing applications into mathematical instruction helped merely to convince the community of mathematics teachers that this evolution was inevitable and to induce them to combat reforms intended to bowdlerize mathematics into a set of useful practical tricks.
LES DEBATS AUTOUR DES APPLICATIONS DES MATHEMATIQUES DANS LES REFORMES DE LENSEIGNEMENT SECONDAIRE ` ` AU DEBUT DU VINGTIEME SIECLE par Philippe NABONNAND
INTRODUCTION Publier des contributions traitant des questions de r forme de lenseignement e des math matiques et pr senter des panoramas d crivant lhistoire et lorganie e e sation de cet enseignement dans diff rents pays gurent parmi les principaux e objectifs de la revue LEnseignement Math matique au moment de sa cr ation e e ` en 1899. De nombreuses contributions a cette revue evoquent donc le contexte international de r forme de tous les ordres denseignement au d but du 20e e e si` cle et se font plus particuli` rement l cho des d bats et des interrogations e e e e ` sur les questions de lenseignement des math matiques. En devenant a partir e de 1908 lorgane ofciel de la Commission internationale de lenseignement math matique (CIEM), la revue organisera et suscitera, au moins jusquen e 1914, la discussion sur les questions de p dagogie, de didactique, de d nition e e des curricula et dorganisation de lenseignement des math matiques. e La plupart des auteurs ecrivant dans LEnseignement Math matique sont e des partisans et m me, pour beaucoup, des propagandistes de lesprit de e ` r forme. Bien entendu, la situation reste tr` s vari e dun pays a lautre et les e e e volont s r formatrices sexpriment diff remment en fonction des conjonctures e e e nationales. N anmoins, les nations se consid rant alors 1 ) comme civilis es e e e sont toutes travers es au d but du 20e si` cle par un puissant courant de r forme e e e e de lenseignement, en particulier de celui des math matiques. e
1 ) En 1914, la Conf rence internationale de lenseignement math matique de Paris r unit e e e des d l gu s dAllemagne, Autriche, Belgique, Danemark, Egypte, Espagne, Etats-Unis, France, ee e Hollande, Hongrie, Iles Britanniques, Italie, Roumanie, Russie, Serbie, Su` de et Suisse. e
` ` APPLICATIONS DES MATHEMATIQUES AU DEBUT DU VINGTIEME SIECLE
231
Partout les m mes probl` mes se posent, et presque dans les m mes termes, e e e ` partout aussi lon envisage a peu pr` s les m mes solutions. Sil y a entre e e les solutions adopt es ici et l` des diff rences, qui tiennent evidemment au e a e g nie propre de chaque nation, il y a plus de ressemblances, plus de points e communs entre elles quon ne serait tent de le supposer au premier abord. 2 ) e [Darboux 1914, 192193]
Une des raisons le plus souvent invoqu es pour justier la n cessit de ces e e e r formes est lexigence de former des techniciens et des ing nieurs pour les e e besoins de lindustrie alors en pleine expansion. Alors que lon assiste depuis ` e la seconde moiti du 19e si` cle a l mergence dun enseignement technique e e et professionnel au niveau secondaire et sup rieur, la question de la formation e math matique dispens e par ces etablissements (et en particulier celle de la e e ` place a accorder aux math matiques appliqu es et aux applications des math e e e matiques) devient alors cruciale. Lenseignement g n ral nest pas epargn e e e par la volont de r forme, ni par la question des applications. Comme le e e souligne Carlo Bourlet 3 ), le d veloppement des sciences et la croissance des e connaissances scientiques sont une autre des raisons majeures avanc es par e les promoteurs des r formes. e
Lenseignement des math matiques, dans nos lyc es, coll` ges et gymnases de e e e tous pays, passe actuellement par ce que daucuns nomment une crise et qui nest, en somme, quune ` vre de croissance, un malaise n de la rapidit e e e m me de l volution du savoir humain. [Bourlet 1910, 373] e e
Pour la plupart des r formateurs, les nouveaux programmes doivent insister e sur lunit de la science. En cons quence, on ne peut plus dans tous les e e ordres denseignement se contenter dun enseignement abstrait mais on doit ` au contraire accorder un r le essentiel a lexp rience pour la pr sentation o e e et la compr hension des th ories math matiques ainsi quaux applications de e e e celles-ci aux autres disciplines math matiques. e ` Le but essentiel de cet article est de tenter de comprendre, a partir des contributions parues dans LEnseignement Math matique, quelle place e occupent les applications des math matiques dans les nouveaux plans d tude e e e si` cle ; les r formateurs plaident-ils pour qui apparaissent au d but du 20 e e e lapparition de nouveaux chapitres de math matiques appliqu es dans les e e
2 ) Gaston Darboux poursuivait son discours avec peut- tre un peu doptimisme en 1914 : e Malgr les apparences, qui sont quelquefois contraires, les nations se rapprochent de plus en e ` plus les unes des autres, elles tendent de plus en plus a former une humanit civilis e, un concert e e ` e ` des peuples dans lequel chacun doit sattacher a ex cuter sa partie de mani` re a concourir a e ` lharmonie de lensemble et au bien de tous. 3
` ) Professeur au Conservatoire national des arts et m tiers (Paris), d l gu francais a la CIEM. e ee e
232
PH. NABONNAND
curricula de math matiques ou pr nent-ils un enseignement des math matiques e o e moins abstrait et plus ouvert sur les applications des math matiques ? e
1. QUEL UN
CONTEXTE ?
CONTEXTE, DIVERSEMENT ANALYSE, DE DEVELOPPEMENT DE LENSEIGNE-
MENT SCIENTIFIQUE DANS LENSEIGNEMENT SECONDAIRE
` La plupart des intervenants saccordent a souligner limportance des evolutions subies par lenseignement secondaire. Chacun reconnat que les progr` s de la science et de lindustrie au 19e si` cle doivent se traduire e e dans les plans de formation et les programmes. Pour certains, il ne sagit ` que dadapter lenseignement secondaire a une nouvelle d nition de la e notion de culture g n rale. La mission de lenseignement secondaire reste, e e pour ces derniers, essentiellement educative. Par contre, pour dautres, les r formes doivent int grer une nouvelle vis e de lenseignement : former des e e e usagers des sciences au service de lindustrie. De plus, les missions peuvent se d cliner diff remment selon les ordres denseignement. Il faut rappeler e e que si lobjectif de lenseignement secondaire, classique ou moderne, est de e dispenser aux el` ves une culture, celui de lenseignement professionnel et ` technique est avant tout de former a un m tier. La question de la formation e scientique se pose donc radicalement de mani` re diff rente selon les ordres e e denseignement : les d bats autour de l volution de la notion de culture e e g n rale et des rapports entre culture classique et scientique concernent e e essentiellement lenseignement secondaire moderne. Les curricula scientiques des etablissements techniques sont la plupart du temps subordonn s aux e exigences de la formation pratique et technique. Ainsi, au congr` s de Bruxelles de la CIEM (1910), A. Matthias 4 ) souligne e le bouleversement que sont en train de connatre les curricula dans son pays. Les sciences, depuis le d but du 20e si` cle, ne sont plus cantonn es a un e e e ` r le marginal dans le plan de formation. Au contraire, dispenser une v ritable o e culture scientique devient une des missions de lenseignement (en particulier secondaire) :
e Le v ritable r le des sciences dans lenseignement moyen a et longtemps e o m connu sous linuence pr pond rante des etudes classiques. Aujourdhui on e e e reconnat leur valeur educative. [CIEM 1910, 387]
4
) Repr sentant du minist` re prussien de linstruction publique. e e
233
De m me, E. Beke 5 ), dans son rapport g n ral [Beke 1914], consid` re e e e e que les causes profondes du mouvement de r forme de lenseignement sont e ` a chercher dans la transformation, survenue au 20e si` cle, des id es sur e e la culture g n rale et dans les efforts de lenseignement secondaire tendant e e ` a suivre la transformation de ces id es , et non pas dans les lacunes de e lenseignement ou linsufsance de ses r sultats. e Pour dautres, le r equilibrage des programmes en faveur de lenseignement e des sciences ne peut quentraner une evolution radicale des objectifs de lenseignement secondaire. Ainsi, Carlo Bourlet, dans sa conf rence de 1910 e sur la p n tration r ciproque des math matiques pures et des math matiques e e e e e appliqu es dans lenseignement secondaire, prone la priorit du r le social de e e o l cole sur celui plus traditionnel de formation des individus. e
Notre r le [celui des enseignants] est terriblement lourd, il est capital, puisquil o sagit de rendre possible et dacc l rer les progr` s de lHumanit tout enti` re. ee e e e Ainsi concu, de ce point de vue g n ral, notre devoir nous apparat sous un e e nouvel aspect. Il ne sagit plus de lindividu, mais de la soci t ; et, lorsque nous ee recherchons la solution dun probl` me denseignement, nous devons choisir une e m thode non pas suivant sa valeur educative pour l l` ve isol , mais uniquement e ee e suivant sa puissance vulgarisatrice pour la masse. [Bourlet 1910, 374]
` Pour Bourlet, la valeur dun enseignement ne se mesure plus seulement a ` laune dune quelconque valeur formatrice mais aussi a son utilit . e
Un enseignement moderne ne saurait se contenter de cultiver les facult s de e lesprit, il doit savoir le meubler de faits, nombreux et pr cis. Nous navons e ` pas a former des philosophes qui vivront en savants ermites, mais des hommes daction qui devront contribuer, pour leur part, au progr` s humain. [Bourlet e 1910, 374]
Dans sa conf rence sur ladaptation de lenseignement secondaire aux e progr` s de la science, Emile Borel, pourtant partisan des r formes 6 ), est nete e tement moins enthousiaste devant les evolutions utilitaristes de lenseignement, quil consid` re cependant comme in vitables. Selon Borel, la conception de e e lenseignement secondaire selon laquelle il sagit avant tout de former des ` e hommes cultiv s, poss dant cette culture g n rale si difcile a d nir doge e e e matiquement, mais dont lid e est cependant fort claire [Borel 1914, 199] e est remise en cause par les partisans dun enseignement strictement utilitaire. Borel, tout en d fendant une evolution lente, sage et prudente souligne e
5 6
` ` ) Professeur a luniversit de Budapest, d l gu hongrois a la CIEM. e ee e ) Borel prononce le 3 mars 1904 une conf rence dans laquelle il pr ne, entre autres, une e o
` orientation plus pratique des exercices de math matiques et d fend une conception a la fois e e th orique et pratique de l ducation math matique [Borel 1904]. e e e
234
PH. NABONNAND
` ` dailleurs les dangers quil y aurait a suivre une tendance trop utilitariste et a suivre de trop pr` s les modes passag` res. e e UNE CRISE DE LENSEIGNEMENT DES MATHEMATIQUES QUI APPELLE DES
REFORMES PROFONDES ` Les math matiques etant reconnues comme linstrument indispensable a e l tude des ph nom` nes naturels et economiques, le r le formateur des math e e e o e matiques est r evalu . Les math maticiens et les enseignants de math matiques e e e e de tous pays accomplissent en quelques ann es une v ritable r volution e e e culturelle. Les nouveaux programmes et les nouvelles instructions insistent sur la n cessit dun enseignement qui sinscrive dans la vie pratique et offre e e de nombreuses applications. L l` ve doit se rendre compte des liens nombreux ee qui existent entre les sciences math matiques et la vie pratique. Il faut agir e sur le d veloppement de la pens e non pas par des connaissances isol es mais e e e par des connaissances qui soient en relation etroite avec lactivit journali` re e e et les id es usuelles. Pour r aliser cet objectif les r formateurs, en g n ral, e e e e e insistent sur quatre points : 1) tenir compte des domaines de la vie pratique, en particulier d velopper e lint r t pour les questions economiques ; ee 2) d velopper la conception de lespace en pr sentant la g om trie de mani` re e e e e e 7 plus intuitive ) et en centrant l tude de la g om trie sur celle de e e e transformation g om trique 8 ) ; e e 3) utiliser les repr sentations graphiques ; e e 4) introduire les el ments du calcul innit simal avec les applications. e Il faut aussi noter que le souci dassurer une formation math matique e plus appliqu e est le signe de la crainte de voir lenseignement des math e e matiques assur directement par les utilisateurs (professeurs de physique, de e technologie, ). Cette inqui tude est particuli` rement sensible en Allemagne. e e Ainsi, H. E. Timerding 9 ) fait etat au congr` s de Milan de la CIEM (1911), e dune grande tendance amath matique ou antimath matique en Allemagne. e e Il donne ainsi lexemple dune ecole foresti` re o` lon a supprim le poste e u e
7 ) Certains r formateurs proposent de pr senter et m me de d montrer exp rimentalement les e e e e e th or` mes de g om trie. e e e e 8 ) Pour plus de pr cisions, on peut consulter larticle de R. Bkouche dans ce volume ou e [Bkouche 1991]. 9 ` ) Professeur a lEcole technique sup rieure de Braunschweig, membre de la Sous-commission e allemande de la CIEM.
235
` de professeur de math matiques et transmis lenseignement math matique a e e des professeurs sortant de la carri` re foresti` re elle-m me [CIEM 1911, 485]. e e e De m me, C. Godfrey 10 ), au congr` s international des math maticiens e e e de Rome, dans sa description du processus de r formes en Angleterre, e insiste sur limportance des utilisateurs des math matiques, en loccurrence les e ing nieurs. En effet, selon lui [Godfrey 1908], malgr la prise de conscience de e e certains enseignants, linstitution ne put se r former elle-m me et limpulsion e e n cessaire vint des ing nieurs . Ces derniers se sont rendu compte de la e e n cessit dune meilleure formation scientique pour leur corps et afrment e e quon ne peut pas savoir trop de math matiques pourvu que ce soit de bonnes e math matiques . Avec la cr ation au sein de luniversit de Cambridge dune e e e section ding nieurs, lenseignement des math matiques sans base pratique e e est en question.
Ce mouvement amena la formation de divers comit s qui compar` rent les e e opinions des hommes du m tier et des matres d cole et trouv` rent que laccord e e e etait possible sur la plupart des points. Les professeurs reconnurent que des sujets utiles pouvaient etre aussi educatifs que les futilit s conventionnelles qui e avaient ni par sidentier avec les math matiques enseign es dans les ecoles. e e De m me que les math matiques sup rieures pures gagnent en valeur et en e e e int r t par un contact plus intime avec les probl` mes pos s par les physiciens et ee e e deviennent en revanche irr elles et sans but quand elles sont s par es de leurs e e e e applications, de m me les math matiques el mentaires ont trouv leur salut e e e dans lintroduction des applications sans nombre fournies par la vie industrielle moderne. [Godfrey 1908, 462]
Les nouveaux programmes darithm tique int` grent ainsi lusage des tables e e ` de logarithmes a 4 d cimales plus pratiques et pouvant etre enseign es d` s e e e l ge de 14 ans . Quelques ecoles ont m me pr vu un cours de travaux a e e e ` exp rimentaux simples dans un laboratoire (pour les el` ves de 13 a 15 ans) et e faisant explicitement partie du cours de math matiques [Godfrey 1908, 464]. e De m me, lenseignement de la g om trie comporte une part exp rimentale, e e e e lenseignement de la trigonom trie est essentiellement num rique (usage de e e tables, r solution de probl` mes pratiques bas s sur des observations faites e e e l` ves avec un th odolite simpli ) [Godfrey 1908, 472]. Quant a la ` par les e e e e m canique, lenseignement de la statique est fond sur un cours exp rimental. e e e Par contre, lenseignement de la dynamique reste th orique car il nest e pas facile dorganiser des travaux exp rimentaux sur ce sujet . Il en r sulte e e e des difcult s pour une grande part des el` ves concernant cette partie. La e conclusion de Godfrey est que la trigonom trie et la statique formeront peute
10
) Directeur du R. N. College, Osborne.
236
PH. NABONNAND
etre pendant un certain temps la limite des etudes math matiques de la plupart e des jeunes gens dune ecole publique. Dans une note de 1909 du Board of Education 11 ) concernant lenseignement de la g om trie dans lenseignement secondaire, il est indiqu qu il est e e e ` maintenant usuel dadjoindre plus ou moins de travail graphique a lalg` bre e e et quainsi on fait entrer un el ment de r alit dans un sujet tr` s abstrait e e e et irr el, ce qui ne peut avoir que de bons r sultats [Bruce 1910, 248]. e e Des exemples de r solution graphique d quations du second degr , et de e e e repr sentation de fonctions du second degr , sont d crits. Un premier int r t e e e ee (pratique) dune telle m thode est dentraner au calcul num rique ainsi qu` e e a e lexactitude dans les mesures et le dessin . Dautre part, des el` ves plus avanc s pourront etendre ces techniques aux equations de degr 3 et pourront e e se rendre compte de la puissance de la m thode quils ont entre les mains . e e Le second int r t (p dagogique) est quavec ces m thodes, les el` ves se ee e e rendront ainsi evidemment matres des notions essentielles de lalg` bre . e Cependant, il ne faut pas sacrier la rigueur de lexpos de lalg` bre, et le e e travail graphique ne doit pas etre un but, mais un moyen . En Allemagne, le courant r formateur, particuli` rement actif, est organis e e e autour de la personnalit de F lix Klein qui depuis fort longtemps pronait une e e modernisation des enseignements de math matiques et une prise en compte e plus importante des math matiques appliqu es 12 ). Les propositions de M ran e e e (1905) et de Stuttgart (1906) d nissent le renforcement de lintuition de e lespace en fondant lenseignement de la g om trie sur la notion de transfore e mation g om trique et le d veloppement de lid e de fonction comme objectifs e e e e principaux de lenseignement des math matiques dans les ecoles sup rieures. e e ` Pour parvenir a ce but, il faut ` 1) ordonner lenseignement de facon a mieux ladapter au d veloppement e naturel de lesprit ; 2) d velopper autant que possible cette facult dobservation math matique e e e des ph nom` nes qui nous entourent par un choix appropri dapplications ; e e e ` ` 3) arriver peu a peu a la conception de lunit de la science en concentrant e tout lenseignement autour de la notion de fonction, aussi bien au point de vue alg brique quau point de vue g om trique. [EM 1911a, 67] e e e
11 ) Bureau du gouvernement britannique qui soccupe des questions denseignement public (au sens francais du terme). Le Board of Education avait pris en charge lorganisation et les travaux de la Sous-commission britannique de la CIEM. 12 ) Sur le mouvement de r forme de lenseignement des math matiques en Allemagne, on e e peut consulter en particulier [Schubring 1989], [Tobies 1989] ou larticle de G. Schubring dans ce volume.
237
Ces propositions sont appliqu es plus ou moins ofciellement dans la e ` plupart des Etats allemands a partir de 1908. De plus, en Prusse, il est accord e aux trois genres d coles (Gymnasien, Realgymnasien et Oberrealschulen) e l galit en ce qui concerne les droits quelles accordent , traduisant ainsi e e institutionnellement un certain r equilibrage entre les formations classique et e ` moderne ainsi quentre enseignement g n ral et enseignement a vocation plus e e technique. La r forme de 1902 en France accorde aux sciences une place beaucoup e plus importante dans les li` res classique et moderne de lenseignement e secondaire. Les nouveaux programmes accordent aux formations scientique et litt raire la m me importance ; la volont afch e par le minist` re est de e e e e e promouvoir une culture scientique au m me titre quune culture litt raire 13 ). e e
Les lettres sont et resteront comme par le pass , des institutrices eprouv es e e quil serait impossible de suppl er dans leur domaine. Mais dans le domaine e qui est celui des sciences positives, on attend des sciences plus deffets que par le pass , pour la formation des esprits. (L. Liard (1904), cit par Beke [1914, e e 245246])
` Pour Gaston Darboux, a qui fut con e la pr sidence de la commission de e e r vision des programmes scientiques, les principaux acquis en math matiques e e de la r forme de 1902 se r sument aux quatre points suivants : e e
1o 2o 3o 4o e lintroduction dans lenseignement el mentaire du Calcul des d riv es et e e m me de notions de Calcul int gral ; e e lemploi syst matique dans la g om trie des m thodes de transformation e e e e qui simplient l tude et apportent un principe de classication ; e le d veloppement donn aux applications qui sont pos es par la pratique, e e e ` a lexclusion de ces probl` mes qui nont aucune racine dans la r alit ; e e e le d veloppement aussi complet que possible de linitiative personnelle chez e e ` tous les el` ves qui prennent part a lenseignement et une pr occupation e incessante dune bonne formation de lesprit. [Darboux 1914, 197]
Partout en Europe 14 ) se pose la question de r former lenseignement des e math matiques dans un sens plus pratique, ce qui se traduit le plus souvent e ` par la volont de fonder a partir de lexp rience quotidienne lenseignement e e de la g om trie et celle dintroduire dans les cursus le calcul diff rentiel et e e e int gral, reconnu comme loutil essentiel des applications des math matiques e e dans les autres sciences.
13 ) Pour plus de pr cisions sur la r forme de 1902 en France, on peut consulter les ouvrages e e de B. Belhoste [1995], N. Hulin [2000], ainsi que [Belhoste, Gispert & Hulin 1996]. 14 ` ) Le mouvement de d veloppement des enseignements de math matiques a la n du 19e e e si` cle naffecte pas lItalie o` il subit un recul, devenant m me optionnel dans certaines li` res e u e e [EM 1912c, 253].
238 2. LES
PH. NABONNAND
APPLICATIONS DANS LES DISCUSSIONS
AUTOUR DE LENSEIGNEMENT DES MATHEMATIQUES Dans sa d claration dintention [CIEM 1908, 445458], la CIEM se donne e comme objectif de faire une enqu te et publier un rapport g n ral sur les e e e tendances actuelles de lenseignement math matique dans les divers pays e [CIEM 1908, 450]. Cet etat des lieux concerne lenseignement de la premi` re e ` initiation a lenseignement sup rieur, et aussi bien lenseignement g n ral que e e e lenseignement technique ou professionnel. M me si elle nest pas cantonn e e e ` a cet ordre denseignement, la question des math matiques appliqu es est e e ressentie par les auteurs de LEnseignement Math matique comme beaucoup e plus cruciale lorsquil sagit des formations techniques, en raison en particulier de leur relative nouveaut et donc dune moindre pesanteur des traditions. e
En raison de limportance croissante que prennent ces ecoles [les ecoles techniques ou professionnelles] et des exigences nouvelles quon ne cesse de montrer vis-` -vis de lenseignement math matique, il y aura lieu daccorder dans cette a e enqu te une large place aux math matiques appliqu es. [CIEM 1908, 452] e e e
QUOI SERVENT LES MATHEMATIQUES DANS UNE FORMATION ?
Les partisans les plus r solus des r formes, tout en insistant sur la e e valeur utilitaire des math matiques, ne c` dent en rien aux opposants ou aux e e ` r ticents 15 ) quant a la valeur educative et disciplinaire de celles-ci. Certes, e les positions sont plus ou moins nuanc es selon les types de formation et le e statut de branche principale ou branche secondaire que les math matiques e y occupent. De plus, les discours sur la mission culturelle de lenseignement concernent essentiellement lenseignement secondaire classique et moderne. ` N anmoins, les r formateurs vont sattacher a montrer que louverture des e e curricula aux applications et la disparition des programmes de math matiques e dun certain nombre de notions et de m thodes, comme certaines virtuosit s e e purement calculatoires de g om trie analytique ou darithm tique, au prot e e e de lintroduction du calcul diff rentiel et int gral ne constituent en rien un e e appauvrissement du contenu du cours de math matiques et encore moins un e affaiblissement de sa valeur formatrice. Bourlet, qui d fend la pr eminence du r le social de lenseignement en e e o g n ral, pr ne bien entendu un enseignement essentiellement utilitariste : e e o
15 ) Les contributeurs de LEnseignement Math matique et les participants aux diverses e commissions de la CIEM sont pour lessentiel des partisans des r formes. Le discours des e opposants napparat donc, dans le corpus que nous nous sommes x , quindirectement, dans e les r ponses des partisans des r formes. e e
239
` e [] il ne nous est plus permis maintenant de pr senter a nos el` ves la science e math matique sous un aspect purement sp culatif et [] il nous faut, co te e e u ` que co te, plus encore pour rendre service a la soci t dans son ensemble, u ee qu` chacun de nos etudiants en particulier, nous efforcer de faire plier les a abstractions math matiques aux n cessit s de la r alit . [Bourlet 1910, 374] e e e e e
Selon Bourlet, le fait majeur au d but du 20e si` cle est la pr dominance de e e e lindustrie parmi les activit s humaines. La n cessit se fait jour de pr parer e e e e ` ` ` les jeunes gens [] a connatre, a pratiquer et a faire progresser les sciences exp rimentales o` cette industrie puise ses forces . Il faut donc ecarter de e u lenseignement des math matiques tout ce qui naura pas une utilit plus ou e e moins directe dans les applications. Cependant, un tel programme ne conduit ` pas n cessairement a une baisse de contenu du cours de math matiques, car e e si l on fait un tableau complet des connaissances strictement indispensables ` a un ing nieur ordinaire, on sapercoit aussitot que le champ ainsi born e e est encore immense . Parmi les evolutions n cessaires, lenseignement de e lanalyse en France a d sadapter aux sciences appliqu es en introduisant u e par exemple la notion de fonction, base de toute etude des ph nom` nes e e naturels, et de sa repr sentation graphique dans des manuels de pr paration e e au baccalaur at. Cette evolution fondamentale du programme danalyse que e constitue lintroduction du calcul diff rentiel dans les programmes des derni` res e e ` ann es du lyc e est due, selon Bourlet, a lapparition denseignements de e e physique et de m canique dans lenseignement secondaire. Il conclut en e afrmant que la limite entre les math matiques pures et les math matiques e e appliqu es nexiste pas, car ces deux sciences, loin d tre s par es, doivent e e e e sans cesse sentraider et se compl ter [Bourlet 1910, 386]. e De m me, la r forme qui sop` re en Allemagne dans les ann es 19081909 e e e e passe par une orientation plus pratique et plus appliqu e des math matiques et e e ` accorde a lintuition et aux applications un role pr pond rant [EM 1910a, 63]. e e En 1912, dans un rapport sur les probl` mes commerciaux et lenseignement e des math matiques dans les ecoles secondaires, Timerding [EM 1912a, 6063] e reprend la discussion de lobjectif et de lutilit de lenseignement des math e e matiques.
` e Tout lenseignement d pend du but que lon assigne a l cole. Les uns veulent e ` que, par une gymnastique intellectuelle intense, elle habitue lesprit a bien penser et craignent toutes les questions pratiques que compliquent trop les contingences de la vie pour quelles soient un bon aliment de la pens e pure. e Les autres, se d ant des esprits trop logiques, d sirent, au contraire, que l cole e e e ` e inculque des connaissances pr cises a ses el` ves et les mette en contact avec e la complexit des choses. [EM 1912a, 60] e
240
PH. NABONNAND
Il ne choisit pas entre ces deux points de vue dont on a vu quils d pendent e essentiellement du public auquel on sadresse ; par contre, il d veloppe une e id e int ressante : un enseignement d lib r ment appliqu comme celui de e e e ee e larithm tique politique (un m lange de th orie des probabilit s appliqu es, e e e e e de math matiques nanci` res, de statistiques, de math matiques appliqu es e e e e aux assurances, ) peut n anmoins satisfaire les exigences culturelles de e lenseignement secondaire g n ral et est loccasion de montrer que les notions e e e math matiques ne sont pas arbitraires, mais quon y a et amen par la force e e des choses . On retrouve le m me type de pr occupation dans un rapport sur lenseie e gnement des math matiques dans les Realschulen su doises. Le but de e e ` lenseignement des math matiques dans ces etablissements a vocation intere m diaire et de caract` re technique doit etre essentiellement pratique, ce e e qui n cessite des m thodes denseignement pratique. Mais, la r solution e e e dun probl` me par une equation ne doit pas exclure syst matiquement le e e raisonnement lorsquil peut etre utile [EM 1911e, 171]. En lisant ces d clarations, on pourrait penser que lopposition entre les e points de vue des tenants et des opposants des r formes est, en fait, des e plus t nues. Pourtant, le foss reste beaucoup plus profond quil napparat e e dans le discours des r formateurs. En effet les opposants aux r formes, en e e ` exprimant leurs inqui tudes relatives au contenu et a laffaiblissement de e la valeur educative du cours de math matiques, salarment surtout de la e perte dautonomie de cet enseignement. Il est vrai quen insistant sur les applications des math matiques et les liens de celles-ci avec les autres sciences, e les r formes remettent en cause larchitecture de la formation math matique e e dans lenseignement secondaire (et sup rieur) et soumettent lenseignement e ` purement disciplinaire des math matiques a celui plus g n ral dune m thode e e e e scientique [Perry 1909, 137]. Ainsi, les nouveaux programmes des math matiques dans les etablissements e ` e secondaires sup rieurs en Allemagne visent explicitement a donner aux el` ves : e
un coup dil scientique sur la parent des sujets math matiques trait s a e e e ` l cole ; une certaine aptitude de la conception math matique et son emploi e e ` a la r solution de probl` mes particuliers ; enn et surtout la p n tration de e e e e limportance des math matiques pour la connaissance exacte de la nature. [EM e 1906, 58]
Avec de tels objectifs, l l` ve est donc cens acqu rir une formation en ee e e math matique non seulement pr cieuse en elle-m me mais qui sera utile e e e ` dans lexercice de sa profession, au moins pour ceux qui se destinent a
241
une carri` re scientique ou technique. On retrouve le m me type de soucis e e p dagogiques dans les nouveaux programmes autrichiens publi s en 1909 : e e
Dans les conditions actuelles des ecoles r ales, lenseignement math matique e e e a pour but la pratique des math matiques el mentaires, y compris la notion de e fonction, comme pr paration aux ecoles sup rieures ; il ne doit pas avoir en e e vue une culture sp ciale, mais contribuer au d veloppement g n ral de lesprit e e e e par la science. [EM 1910b, 338]
Pour ce qui int resse les math matiques appliqu es et plus g n ralement e e e e e la notion dapplication des math matiques, ce programme se traduit par : e 1) une simplication du champ d tude par la liaison des branches ayant des e relations les unes avec les autres, et 2) une adaptation du programme de math matiques aux branches correspone dantes et aux applications de la vie r elle. e QUELLE
RIGUEUR POUR UN ENSEIGNEMENT PRATIQUE ?
Une des principales inqui tudes quant aux r formes est quen souvrant e e aux applications, lenseignement des math matiques perde toute rigueur. e Traditionnellement, le r le formateur essentiel des math matiques est justement o e e de faire acqu rir aux el` ves la m thode logique et lesprit de rigueur. Ceux qui e e d fendent un enseignement fond sur les applications consid` rent que le r le e e e o formateur des math matiques r side aussi dans lacquisition dune m thode e e e scientique dans laquelle les math matiques ont toute leur place. e A. N. Whitehead [1913, 105113], qui evoque cette crainte dune perte de la rigueur dans lenseignement math matique, distingue deux cat gories e e e de public dans lenseignement el mentaire : ceux qui d sirent limiter leur e formation math matique et ceux qui au contraire ont besoin dune education e math matique plus cons quente pour leur carri` re professionnelle future. Pour e e e les premiers, selon Whitehead, lenseignement math matique, m me sil doit e e ` e rester a un niveau el mentaire, doit viser deux objectifs : d velopper la e facult dabstraction et d velopper la facult de raisonnement logique. Aussi e e e lenseignement des math matiques doit-il etre dune rigueur logique sans e concession. Cependant, il ne faut pas oublier que la pr cision logique est un but e et non le point de d part de lenseignement, et donc celle-ci doit etre obtenue e par approximations successives . Concernant la seconde cat gorie d l` ves, e ee ` il consid` re comme une erreur profonde lopinion (largement majoritaire a e ses yeux) quil soit possible denseigner les math matiques avanc es du seul e e point de vue de lutilit pour les physiciens ou ing nieurs sans sint resser e e e ` a la logique et la th orie. Il est important pour les physiciens et ing nieurs e e davoir un esprit entran math matiquement (mathematically trained mind) et e e
242
PH. NABONNAND
on ne peut pas se contenter dune connaissance quasiment m canique en vue e des applications. Whitehead conc` de que lon peut introduire les notations et e les premi` res notions du seul point de vue des applications mais l ducation e e ` math matique des scientiques appliqu s doit consister a rendre ces notions e e ` pr cises et a donner des d monstrations pr cises. Veronese d fend une position e e e e analogue au congr` s de Milan de la CIEM (1911). Une des questions soumises e ` a la discussion du congr` s etait : Dans quelle mesure peut-on tenir compte, e dans les ecoles moyennes (lyc es, coll` ges, gymnase, ecoles r ales, etc.), e e e de lexpos syst matique des math matiques ? La discussion tourne autour e e e dune opposition entre un enseignement d ductif, suppos rigoureux, et un e e enseignement intuitif exp rimental, plus laxiste. Pour Veronese en raison du e r le educatif des math matiques, le r le dun enseignement de type intuitif o e o ` e et exp rimental doit etre r duit a pr parer un enseignement ax sur la seule e e e ` d duction, au moins pour les ecoles qui pr parent a lenseignement sup rieur. e e e En particulier,
si lindustrialisme ou lutilitarisme mat riel avait [] des inuences pr pond e e e rantes dans lenseignement des ecoles moyennes, les math maticiens devraient e les combattre. 16 ) [CIEM 1911, 465]
M me Bourlet, dont on a vu lengagement militant pour un enseignement e plus pratique et plus appliqu des math matiques, afrme quil ne faut rien e e sacrier des qualit s de rigueur, de logique et de pr cision qui sont lapanage e e ` des math matiques . De m me, Timerding, qui ne veut surtout pas aller a e e lencontre de la tendance exp rimentale qui se d veloppe dans lenseignement e e de la physique en Allemagne et qui prone un enseignement des math matiques e ` qui, sans renoncer a ses buts propres, doit tenir compte de la r alit et des e e applications , a le souci non seulement de la rigueur de cet enseignement mais aussi que ces exigences ne soient pas contredites lors des applications :
` On ne doit pas, ici, rappeler l l` ve a la rigueur et a lexactitude, alors que l` un ee ` a laisser-aller commode dans lexpression et le raisonnement est non seulement permis, mais encore donn en exemple. [EM 1911b, 6970] e
Pourtant, dautres arguments se font entendre ; selon le m me Timerding, les e ` objectifs que lon assigne a lenseignement math matique interrogent aussi la e m thode de cet enseignement. En particulier, il se demande sil faut observer e e toute la rigueur m me dans un enseignement el mentaire o` lon ne veut traiter e u que les premiers principes dans un but pratique ou au contraire si lon ne peut pas se servir dune induction partielle au lieu de la d duction pure pour e
16 ) Veronese nuance l g` rement sa position en soulignant quil ne faut pas faire de la rigueur e e excessive dans lenseignement moyen .
243
` faciliter les etudes ou m me avoir recours a des m thodes exp rimentales e e e [CIEM 1911, 495]. De la m me mani` re, G. Scorza 17 ), auteur dun rapport sur lenseignement e e des math matiques dans les ecoles et instituts techniques italiens (enseignement e secondaire), regrette les pr occupations excessives de prudence rigoriste e qui compliquent inutilement lapprentissage [EM 1912e, 420]. Au congr` s e de Milan, J. W. A. Young 18 ) accorde, lui aussi, une importance capitale aux approches de type pratique et consid` re que e
les d buts devraient se faire dune facon concr` te, aussi bien dans lenseignement e e secondaire en g n ral que dans le travail dune ann e particuli` re ou dans e e e e lexposition dun sujet sp cial quelconque ; les proc d s abstraits (abstraits ree e e ` lativement a la maturit et au degr davancement de l l` ve) napparaissant que e e ee pour eviter de trop nombreuses r p titions concernant des exemples concrets e e essentiellement pareils. [Young 1911, 474]
Selon Young, le but de lenseignement de la classe nest pas de faire des math matiques abstraites, mais plut t des math matiques pr sentant par-ci e o e e par-l` des proc d s abstraits. a e e QUELLE PLACE POUR LES APPLICATIONS DANS UN COURS DE MATHEMATIQUES ?
La question du contenu dun cours de math matiques appliqu es dans e e lenseignement secondaire nest jamais abord e explicitement dans les d bats e e de LEnseignement Math matique. Une des raisons est quil ny a pas daccord e sur la formation des utilisateurs de math matiques : doit-on dispenser un cours e de math matiques g n rales centr autour dun corpus de notions consid r es e e e e ee comme utiles ou, au contraire, un cours centr sur les seules applications e pratiques ? Une solution est de tenter de concilier les deux objectifs :
Il faut que ce programme renferme des questions dordre r ellement pratique e ` et ne soit pas r duit a une pure gymnastique c r brale, ce qui ne veut pas e ee dire toutefois que le cours de math matiques appliqu es soit transform en un e e e cours de physique exp rimentale. e Un programme bien compris, qui initierait les auditeurs aux m thodes fondae mentales de la physique et leur fournirait en m me temps des r sultats de nature e e math matique en evitant cependant de trop grandes difcult s analytiques, e e constituerait une excellente base daction commune pour le math maticien et e le physicien. [EM 1912b, 73]
17 18
` ` ) Professeur a lInstitut technique de Palerme, d l gu italien a la CIEM. ee e ` ` ) Professeur a lUniversit de Chicago, d l gu des Etats-Unis a la CIEM. e ee e
244
PH. NABONNAND
De m me, Fehr soutient une position m diane et afrme qu` c t des e e a oe cours destin s aux math maticiens et physiciens, il est n cessaire quil y e e e ait un cours dit de math matiques g n rales portant sur les notions les plus e e e utiles . Il souligne que dailleurs les probl` mes commencent avec la d nition e e de ce que sont ces notions. P dagogiquement, il insiste sur la n cessit de e e e ` e travaux pratiques qui doivent montrer a l tudiant, mieux quon ne peut le faire par des exemples dans un cours g n ral, comment les math matiques e e e interviennent r ellement dans les applications . Il ajoute : e
Il est d sirable que les ecoles sup rieures apportent une attention toute sp ciale e e e au d veloppement de cet enseignement pratique pour en faire un v ritable e e laboratoire math matique. [CIEM 1911, 494] e
De plus, il faut distinguer les formations pour lesquelles les math matiques e sont une mati` re principale et celles o` elles sont consid r es plut t comme une e u ee o ` branche accessoire, destin e a abr ger certains raisonnements et a formuler e ` e dune facon particuli` rement br` ve tout un ensemble de r sultats . Ainsi, e e e P. Rollet 19 ), dans son rapport sur lenseignement technique secondaire en France, pr sente lenseignement math matique dans ce type de formation e e comme n tant ni une n, ni un but (au contraire des formations g n ralistes e e e dispens es dans les coll` ges et lyc es). Il y a donc lieu d carter toutes e e e e ` m thodes et d monstrations qui ne concourent pas a la n cherch e ou e e e ` au but poursuivi , a savoir former des ouvriers, des contrematres ou des techniciens. Il est par contre vivement conseill dinsister sur les liens avec e les cours techniques et les applications.
Acceptant linuence du milieu technique dans lequel ils vivent, les professeurs de math matiques ont su caract riser nettement leur enseignement et lui donner e e son adaptation pratique, tout en ne perdant pas de vue le role educatif qui reste le propre des math matiques. [EM 1912d, 325] e
Selon H. Gr nbaum 20 ), dans un rapport sur lenseignement math matique u e en Allemagne dans les ecoles techniques moyennes pour lindustrie m canique, e ces derni` res assignent aux math matiques le r le de science accessoire (par e e o ` opposition a celui de science educative dans les lyc es), destin e a r soudre e e ` e des probl` mes techniques. Dans ce type de formation, les applications sont le e ` but supr me a poursuivre, et lenseignement des math matiques, des sciences e e naturelles et des branches techniques sy fait simultan ment. Les r sultats e e math matiques principaux doivent etre enonc s et d montr s, en ecartant e e e e
19 ` ) Directeur de lEcole municipale professionnelle Diderot a Paris, membre de la Souscommission francaise de la CIEM. 20
) Collaborateur de la Sous-commission allemande de la CIEM.
245
syst matiquement tous les sujets qui nont pas dapplications techniques, tels e que la trigonom trie sph rique, la g om trie synth tique, etc. Le choix des sue e e e e jets trait s dans le cours doit se conformer au crit` re de lapplication pratique : e e
A ce titre, les calculs les plus simples, les constructions g om triques les plus e e e el mentaires doivent etre exerc s aussi bien que les parties soi-disant sup rieures e e e des math matiques, tels que les el ments du calcul innit simal, dont lemploi e e est courant dans les publications techniques. [EM 1911c, 156]
La g om trie descriptive, la m canique et les m thodes graphiques constituent e e e e le programme de math matiques appliqu es. e e De m me, les nouveaux programmes autrichiens des ecoles techniques e insistent sur limportance du calcul num rique et de la notion dapproximation, e sur la possibilit de pr senter graphiquement la fonction logarithmique et sur e e lutilisation des tables num riques. Les instructions insistent sur limportance e des exercices et des probl` mes qui doivent toucher aux diff rentes branches e e de lenseignement et pr senter des rapports avec la vie courante [EM 1910b, e 330332]. Enn, il faut tenir compte des traditions souvent nationales qui inuent sur la conception g n rale du cours de math matiques. Dans la conclusion e e e de sa conf rence d j` cit e, Bourlet [1910] pr cise que lenseignement en e ea e e e ` France est traditionnellement g n raliste et oblige les el` ves a recevoir une e e ` instruction g n rale tr` s etendue . Il poursuit en soulignant quen France, a la e e e diff rence de lAllemagne, on ne concevrait pas un cours de math matiques e e uniquement pour des chimistes fait dans lesprit de la sp cialisation etroite . e Il ne faut pas n gliger non plus que la question dun enseignement de e ` math matiques appliqu es est a la fois beaucoup plus cruciale et plus simple e e ` e a r soudre pour les formations sup rieures ding nieurs ou de physiciens que e e dans lenseignement secondaire m me technique. En effet, personne ne remet e en cause la n cessit dun enseignement de math matiques pour ces formations e e e et tout le monde saccorde avec les id es expos es par E. Poss dans son e e e rapport sur lenseignement technique sup rieur en Russie : e
L tude des sciences math matiques nest pas le but principal des ing nieurs, e e e mais elle leur est indispensable comme etude auxiliaire, les Math matiques e etant la base de toutes les sciences techniques pr cises. [EM 1911f, 337] e
Dans les formations techniques sup rieures, le cours de math matiques e e appliqu es apparat souvent comme un compl ment dune formation g n rale e e e e ant c dente. La formation des techniciens et ing nieurs, selon Jouglet 21 ) e e e dans une conf rence sur lorganisation de lenseignement technique dans e
21
) Ing nieur des arts et m tiers. e e
246
PH. NABONNAND
les ecoles darts et m tiers [CIEM 1910, 401], doit comporter trois volets : e a) la culture g n rale, b) la culture scientique et c) la culture technique 22 ). e e Dans cette perspective, lenseignement math matique est d ni comme l tude e e e ` de compl ments de math matiques n cessaires a la poursuite des questions e e e de m canique et d lectricit . De la m me mani` re, E. Czuber 23 ), dans sa e e e e e ` e description de lenseignement des math matiques pures a l cole technique e ` sup rieure de Vienne [EM 1911d, 164166], assigne a cet enseignement e (calcul innit simal et g om trie) de servir de base aux autres branches, e e e ` e de permettre a l tudiant la lecture des livres techniques et de lui donner une ind pendance math matique sufsante. Cet enseignement est compl t par un e e ee ` cours de math matiques appliqu es ; a cet egard, il faut signaler un exemple e e int ressant concernant cette ecole, celui du cours de technique dassurance e qui comprend, outre des conf rences de math matiques pures, le calcul des e e probabilit s, la statistique math matique et les math matiques des assurances. e e e Concernant lenseignement secondaire, les d bats autour de lid e dune e e formation math matique appliqu e, ou au moins tenant compte des applicae e tions, se polariseront souvent autour de lintroduction de la notion de fonction et de linitiation au calcul diff rentiel. e Bourlet pr cise que lenseignement des lyc es suft pour donner aux e e jeunes gens les connaissances math matiques (g om trie pure, analytique et e e e descriptive, trigonom trie, alg` bre, calcul diff rentiel et int gral) dont on peut e e e e avoir besoin dans le commerce, l conomie politique et m me les constructions e e civiles et larchitecture. Poske 24 ), dans une conf rence sur lenseignement de la physique [CIEM e 1910, 392], pose tr` s nettement la question de la formation math matique des e e physiciens et afrme que les notions fondamentales du calcul innit simal e doivent etre fournies par lenseignement math matique. De m me, la conclue e sion la plus importante du rapport de Timerding sur les math matiques dans e les trait s de Physique est e
quil est urgent dintroduire les notions de d riv e et dint grale dans le e e e programme de math matiques des coll` ges, et cela assez t t pour quelles e e o puissent etre utilis es et appliqu es concr` tement dans les lecons de physique e e e des classes sup rieures. [EM 1911b, 71] e
22 ) Jouglet pr sente lorganisation des enseignements dans les ecoles des arts et m tiers. e e Ces ecoles constituent laristocratie des ecoles techniques en France et revendiquent pour leurs meilleurs dipl m s le titre ding nieur. La pr sentation des curricula en termes de culture est o e e e caract ristique dune volont de se d marquer des autres etablissements denseignement moins e e e prestigieux, qui ont moins de revendications culturelles. 23 ` ` ) Professeur a lEcole technique sup rieure de Vienne, d l gu autrichien a la CIEM. e ee e 24
` ) Professeur a Berlin, membre de la Sous-commission allemande de la CIEM.
247
Les m mes arguments sont utilis s par le rapporteur sur lenseignement e e ` ` math matique en Su` de, qui souligne que le plan d tudes tend a donner a la e e e notion de fonction la place de notion centrale fondamentale et qu il semble que lexp rience ait d montr que la notion dint grale elle-m me peut etre e e e e e e enseign e a des el` ves de capacit s moyennes et quelle peut etre pour eux e ` e dun grand int r t et dune r elle utilit [EM 1911e, 341]. En particulier, ee e e lintroduction des notions fondamentales du calcul innit simal permet aux e e el` ves qui se destinent aux etudes techniques sup rieures, daborder celles-ci e avec plus de facilit . e
CONCLUSION Malgr la diversit de leurs discours, les r formateurs expriment de mani` re e e e e quasi unanime lexigence de prendre en compte, dans lenseignement et en particulier dans celui des math matiques, les evolutions de la soci t et e ee des demandes sociales vis-` -vis de lenseignement. Cette exigence devrait, a selon eux, se traduire dans les programmes par une prise en compte plus ` importante des applications des math matiques, a la fois parce quil faut former e professionnellement des ouvriers sp cialis s, des techniciens ou des ing nieurs, e e e mais aussi parce quil faut former des individus qui vont vivre en contact avec ` un monde de plus en plus industrialis . De mani` re surprenante, a lexception e e de quelques m thodes graphiques ou num riques, lesprit de r forme ne suscite e e e pas d volution radicale des programmes vers les applications, ni m me e e daccord ou de r exion pour en d gager. La volont de faire evoluer les e e e curricula de math matiques dans un sens plus pratique sexprime certes par la e prise en compte des besoins math matiques des autres disciplines scientiques, e mais lintroduction du calcul diff rentiel et int gral d` s lenseignement e e e secondaire sup rieur est la marque essentielle de l volution des esprits au e e sujet du r le formateur des math matiques. o e Si lon ne peut nier la volont chez les r formateurs de promouvoir un e e enseignement plus pratique des math matiques, on peut n anmoins penser qu` e e a leurs yeux il est beaucoup plus important dintroduire le calcul diff rentiel e et int gral dans les programmes. Largument de la prise en compte des e ` ` applications des math matiques sert a la fois a convaincre les communaut s des e e ` e enseignants de math matiques de lin luctabilit de ces evolutions, a d fendre e e e ` la place des math matiques dans les curricula et a combattre les tentations de e faire enseigner les math matiques sous forme de recettes par les utilisateurs e que peuvent etre les enseignants de physique ou de disciplines techniques.
248
PH. NABONNAND
BIBLIOGRAPHIE BEKE, E. Les r sultats obtenus dans lintroduction du calcul diff rentiel et int gral e e e dans les classes sup rieures des etablissements secondaires. LEnseign. Math. 16 e (1914), 245283. BELHOSTE, B. Les sciences dans lenseignement secondaire francais : textes ofciels. Tome 1 (17891914). Editions Economica/INRP, Paris, 1995. BELHOSTE, B., H. GISPERT et N. HULIN (eds). Les sciences au lyc e. Editions e Vuibert/INRP, Paris, 1996. BKOUCHE, R. Variations autour de la r forme de 1902/1905. In : H. Gispert, La France e math matique. Cahiers dhistoire et de philosophie des sciences 34 (1991), 181 e 213. BOREL, E. Les exercices pratiques de math matiques dans lenseignement secondaire. e Revue g n rale des sciences pures et appliqu es 15 (1904), 431440. e e e Ladaptation de lenseignement secondaire aux progr` s de la science. LEnseign. e Math. 16 (1914), 198210. BOURLET, C. La p n tration r ciproque des math matiques pures et des math matiques e e e e e appliqu es dans lenseignement secondaire. LEnseign. Math. 12 (1910), 372387. e BRUCE, W.-N. Enseignement de la g om trie et de lalg` bre graphique dans les ecoles e e e secondaires [en Angleterre]. LEnseign. Math. 12 (1910), 238253. CIEM. Rapport pr liminaire sur lorganisation de la Commission et le plan g n ral de e e e ses travaux. LEnseign. Math. 10 (1908), 445458. Compte rendu du congr` s de Bruxelles de la Commission internationale de e lenseignement math matique. LEnseign. Math. 12 (1910), 353415. e Compte rendu du congr` s de Milan de la Commission internationale de e lenseignement math matique. LEnseign. Math. 13 (1911), 437511. e Compte rendu du congr` s de Cambridge de la Commission internationale de e lenseignement math matique. LEnseign. Math. 14 (1912), 441537. e ` DARBOUX, G. Discours a la r union douverture de la Conf rence internationale de e e lenseignement math matique (Paris, 14 avril 1914). LEnseign. Math. 16 (1914), e 192197. EM. Articles anonymes de LEnseignement Math matique cit s : e e [1906] Rapport sur lenseignement des math matiques dans les etablissements e ` secondaires sup rieurs a neuf classes [Allemagne]. LEnseign. Math. 8 (1906), e 5777. [1910a] La r organisation de lenseignement math matique dans les ecoles e e sup rieures de jeunes lles en Prusse. LEnseign. Math. 12 (1910), 5970. e [1910b] Les ecoles r ales en Autriche. LEnseign. Math. 12 (1910), 326341. e [1911a] Les ecoles secondaires sup rieures de garcons en Prusse. LEnseign. e Math. 13 (1911), 6267. [1911b] Les math matiques dans les trait s de Physique [Allemagne]. LEnseign. e e Math. 13 (1911), 6971. [1911c] Lenseignement math matique dans les ecoles techniques moyennes pour e lindustrie m canique [Allemagne]. LEnseign. Math. 13 (1911), 154157. e [1911d] Les ecoles techniques sup rieures [Autriche]. LEnseign. Math. 13 (1911), e 164166.
249
[1911e] Ecoles r ales ; gymnases [Su` de]. LEnseign. Math. 13 (1911), 170171 ; e e 339342. [1911f] Universit s et ecoles techniques sup rieures [Russie]. LEnseign. Math. e e 13 (1911), 335338. [1912a] Les probl` mes commerciaux et lenseignement des math matiques dans e e les ecoles secondaires [Allemagne]. LEnseign. Math. 14 (1912), 6063. [1912b] Les relations entre les math matiques et la physique [ Britanniques]. e Iles LEnseign. Math. 14 (1912), 7073. [1912c] Lenseignement math matique dans les Ecoles classiques [Italie]. e LEnseign. Math. 14 (1912), 249253. [1912d] Sur lensemble des etablissements dans lesquels se donne, en France, un enseignement math matique. LEnseign. Math. 14 (1912), 315325. e [1912e] Lenseignement des math matiques dans les ecoles et les instituts e techniques [Italie]. LEnseign. Math. 14 (1912), 416420. FOX, R. and G. WEISZ (eds). The Organization of Science and Technology in France 18081914. Cambridge University Press/Editions de la Maison des Sciences de lHomme, Cambridge/Paris, 1980. GISPERT, H. La France math matique. Cahiers dhistoire et de philosophie des sciences e 34 (1991), 11180. R seaux math matiques en France dans les d buts de la troisi` me r publique. e e e e e Archives internationales dhistoire des sciences 49 (1999), 122149. GODFREY, C. The teaching of mathematics in English public schools for boys. In : Comptes rendus du 4e Congr` s international des math maticiens (Rome, 1908), e e III, 449464 ; Accademia dei Lincei, Rome, 1909. (Trad. fr. LEnseign. Math. 10 (1908), 459474.) HULIN, N. ( d.) Physique et Humanit s scientiques : autour de la r forme de lenseie e e gnement de 1902. Presses Universitaires du Septentrion, Villeneuve dAscq, 2000. e LORIA, G. Sur lenseignement des math matiques el mentaires en Italie. (Extrait e dune communication pr sent e au 3e Congr` s international des math maticiens e e e e ` a Heidelberg.) LEnseign. Math. 7 (1905), 1120. PERRY, J. Lenseignement math matique dans ses rapports avec lenseignement des e sciences. LEnseign. Math. 11 (1909), 136139. POINCARE, H. Les d nitions g n rales en math matiques. LEnseign. Math. 6 (1904), e e e e 257283. ROWE, D. E. and J. MCCLEARY (eds). The History of Modern Mathematics. Vol. II : Institutions and Applications. Academic Press, San Diego, 1989. SCHUBRING, G. Pure and applied mathematics in divergent institutional settings in Germany : the role and impact of Felix Klein. In : [Rowe & McCleary 1989], 171220. TOBIES, R. On the Contribution of Mathematical Societies to Promoting Applications of Mathematics in Germany. In : [Rowe & McCleary 1989], 223248. WHITEHEAD, A. N. The principles of mathematics in relation to elementary teaching. (Conf rence pr sent e au 5e Congr` s international des math maticiens e e e e e Cambridge, 1912.) LEnseign. Math. 15 (1913), 105112. YOUNG, J. W. A. Rapport de M. Young. LEnseign. Math. 13 (1911), 471481.
APPLICATIONS : LES MATHEMATIQUES COMME DISCIPLINE DE SERVICE DANS LES ANNEES 19501960 Applications : mathematics as a service subject in the fties and sixties by H l` ne GISPERT ee
The fties and sixties saw the rise, all over the world, of the new maths reform movement. Considering the harsh criticisms of these different reforms which were expressed as early as the seventies, we might wonder whether the reformers had even thought of mathematical applications. An examination of the journal LEnseignement Math matique has in fact been surprisingly rich in this respect. From the beginning e of the fties till the end of the sixties, ICMI and the reformers paid particular attention to the applications of mathematics, some of the ICMI activities and numerous initiatives being reported in LEnseignement Math matique. e The rst part of this paper deals with two studies carried out by ICMI. The rst one, launched in 1952, had its origin in the idea that mathematical instruction at a given time is intimately linked to the function that mathematics and mathematicians fulll at that time. The reports stressed the fact that the post-war period, which appeared dramatically new and obviously marked by the development of applications, entailed the necessity of a reform of teacher-training taking applied mathematics into account. The second study, ten years later, underlined the little attention paid to the teaching of mathematics applications in all countries except Scandinavia. We shall see what were, then, the positions and priorities of ICMI : at the end of this period (1967) a colloquium was held to answer the question how to teach mathematics so as to be useful, which is not the same thing as how to teach useful mathematics. In the second part of the paper, I study the reections which were made by other organizations engaged in renewing mathematical instruction, in particular within the framework of OECD. Several leading gures, fundamentally committed and representative of the new maths reform movement (especially in France and in Belgium), insisted on the necessary integration of applied mathematics in new mathematical curricula, both for its topics and its methods.
252
H. GISPERT
Lastly, in a third part, I consider the apparent contradiction between the inescapable reference to Bourbaki, which stuck to the new maths movement, and the strong concern of reformers about the applications of mathematics. We may also ask : what were the elds of application concerned when the reformers stressed the universal mathematical language and structures to the prejudice of traditional calculus or geometry ? In fact, what was involved there, was a weakening of the bonds between previously associated disciplines, which had not been replaced by new connections.
APPLICATIONS : LES MATHEMATIQUES COMME DISCIPLINE DE SERVICE DANS LES ANNEES 19501960 par H l` ne GISPERT ee
Cet expos , comme cest la r` gle pour ce colloque, se doit de combiner un e e ` th` me et une epoque en se pliant a une contrainte, celle de sappuyer sur la e revue LEnseignement Math matique et ce quelle a publi des r exions et des e e e d bats en partie impuls s par la Commission internationale de lenseignement e e math matique (CIEM). Cette r` gle, dans mon cas, pourrait paratre paradoxale. e e e En premier lieu, la p riode des ann es 19501960 a et celle de la mont e, e e e partout dans le monde, du mouvement pour la r forme des math matiques e e modernes. Au vu des critiques et bilans des diff rentes r formes, ass n s d` s e e e e e les ann es 1970, on pourrait effectivement se demander sil y avait une place e pour les applications des math matiques dans les r exions des r formateurs. e e e En second lieu, toujours dans ces ann es, la revue LEnseignement Math e e matique etait devenue avant tout une revue de math matiques. Elle avait entam e e une seconde s rie en 1955, apr` s la parution de deux tomes pour lensemble e e des ann es 19421954 (19421950 pour le premier et 19511954 pour le e second) qui terminaient la premi` re s rie inaugur e en 1899. Que pouvait-on e e e bien trouver concernant lenseignement des math matiques, et plus encore des e math matiques comme discipline de service, dans les num ros de ces deux e e d cennies ? e En fait, et cest en soi un premier r sultat int ressant, le d pouillement fut e e e etonnamment fructueux ; la CIEM, dont la vie et les nombreuses initiatives de e` cette p riode ont et a quelques exceptions pr` s rapport es dans la revue, fut d` s e e e e le d but des ann es 1950 et jusquen 1970 pr occup e par les applications des e e e e math matiques comme l taient dailleurs les r formateurs pour lenseignement e e e des math matiques modernes, dont ce fut un souci premier. Cet int r t de la e ee ` organiser mon propos comme suit. revue ma conduit a
254
H. GISPERT
Une premi` re partie sera consacr e a deux enqu tes de la CIEM. La e e ` e premi` re, intitul e Le r le des math matiques et du math maticien dans e e o e e la vie contemporaine , est une des deux premi` res enqu tes lanc es par la e e e nouvelle CIEM en 1952 ; ses r sultats sont publi s dans la revue en 1955. e e Lautre enqu te, qui date du d but des ann es 1960, est pr sent e dans la revue e e e e e en 1964 avec un titre en anglais : Which subjects in modern mathematics and which applications in modern mathematics can nd a place in programs of secondary school instruction ? . Dans une seconde partie, j tudierai les r exions men es dans le cadre e e e dautres organismes que la CIEM mais dune certaine facon en coop ration e avec elle impliqu s dans la r novation de lenseignement des math e e e matiques et en particulier dans le cadre de lOrganisation de coop ration et de e d veloppement economiques (OCDE). Enn, jexaminerai dans la troisi` me e e ` partie lapparente contradiction entre lincontournable r f rence a Bourbaki, ee qui accompagne le mouvement de r forme des math matiques modernes, et e e ce souci appuy des applications des math matiques quont manifest les e e e r formateurs. e
1. LES MATHEMATIQUES MODERNES COMME DISCIPLINE DE SERVICE ` LENSEIGNEMENT MATHEMATIQUE : DUNE ENQUETE A LAUTRE AMSTERDAM, 1954 :
DANS
LE ROLE DES MATHEMATIQUES ET DU MATHEMATICIEN
DANS LA VIE CONTEMPORAINE 1 )
A lorigine de cette premi` re enqu te on trouve une id e afrm e d` s les e e e e e ` premi` res lignes du rapport que G. Kurepa pr sente en 1954 a Amsterdam e e au Congr` s international des math maticiens 2 ) et qui, je pense, garde toute e e ` son actualit aujourdhui. Lenseignement des math matiques a une epoque e e donn e, ecrit-il, est intimement li avec le r le que les math matiques et les e e o e ` e math maticiens jouent a l poque en question. Il sagit ici dune question cl e e pour la CIEM qui devrait permettre dexaminer le probl` me de l ducation e e math matique prise comme un tout . e
1 ` ) Les documents relatifs a cette enqu te sont publi s dans le tome de 1955 de LEnseignement e e Math matique, premier tome de la deuxi` me s rie. Dans le dernier tome de la premi` re s rie, e e e e e celui correspondant aux ann es 19511954, on trouve le compte rendu des d cisions prises par e e le Comit ex cutif de la CIEM pour la pr paration du Congr` s international des math maticiens e e e e e ` dAmsterdam o` gure lannonce et la pr sentation de cette enqu te [pages 72 a 75]. u e e 2 ) Rapport g n ral [Kurepa 1955], suivi en annexe du texte de lancement de lenqu te et e e e du questionnaire associ . e
LES MATHEMATIQUES COMME DISCIPLINE DE SERVICE
255
Six sous-comit s nationaux de la CIEM avaient pris une part active e ` a lenqu te au moment du congr` s. Les d l gu s des sous-commissions e e ee e e nationales avaient en effet et invit s a consulter des personnes quali es e ` e ` dans tous les domaines de la vie contemporaine et a ne n gliger aucun e des aspects du probl` me, dans lordre social, dans lordre intellectuel, dans e lordre scientique, dans lordre des applications pratiques . Lensemble de e ` leurs rapports ont et publi s dans la revue a la suite du rapport de synth` se e e que Kurepa pr sente au congr` s 3 ). Sans revenir sur lensemble des contenus de e e ces rapports, qui d bordent bien evidemment du seul th` me des applications, je e e ` e mattacherai ici a d gager quelques-unes des id es fortes ayant effectivement e ` trait a notre th` me. e Tout dabord, la r exion qua souhait lancer la CIEM est pr sent e e e e e comme n cessaire dans la mesure o` cette p riode de lapr` s-guerre est e u e e d nie comme radicalement nouvelle et ce pour deux raisons. On assiste e ` ` en effet a la fois a des acquisitions vraiment r volutionnaires dans le e ` savoir math matique et a la cr ation de laboratoires de math matiques dans des e e e entreprises economiques, industrielles, commerciales, etc., fait sans pr c dent e e dans lhistoire de lhumanit [Kurepa 1955, 93 et 98]. e Si le premier des deux arguments avait d j` et avanc cinquante ans ea e e plus t t par Carlo Bourlet [1910], le second est nouveau et sp cique de cette o e p riode. Le rapport de synth` se et diff rents rapports nationaux (dont ceux des e e e U. S. A. et des Pays-Bas) soulignent en effet linuence des ann es de guerre e dans cette evolution. Lapr` s-guerre est manifestement une p riode nouvelle, e e ` marqu e par le d veloppement des applications. Cette nouveaut touche a la e e e fois le savoir math matique lui-m me, avec le d veloppement de nouvelles e e e branches issues des recherches de guerre, et la place des math matiques dans e 4 la soci t ). ee Ainsi, la cr ation de chaires de math matiques appliqu es dans les e e e universit s, le d veloppement du nombre de math maticiens travaillant hors e e e du champ acad mique comme aux U. S. A. o` leur proportion parmi les e u math maticiens augmente tr` s rapidement dans ces ann es : plus du quart des e e e deux cents etudiants ayant obtenu leur doctorat en 1951 travaillent au moment de lenqu te dans le monde industriel ou pour le gouvernement. Toujours aux e Etats-Unis, on emploie en 1952 cinquante fois plus de statisticiens quen 1945 dans le domaine du contr le de qualit . Le rapport de Kurepa note la cr ation o e e
3 ) Il sagit des rapports de E. Kamke (Allemagne), O. Weinberger (Autriche), G. Darmois (France), D. van Dantzig (Pays-Bas), G. Ascoli (Italie) et A. M. Gleason (U. S. A.). 4 ` ) Voir a ce propos l tude dAmy Dahan [1996] sur lessor des math matiques appliqu es e e e aux U. S. A. et limpact de la deuxi` me guerre mondiale. e
256
H. GISPERT
de nombreuses nouvelles soci t s savantes qui ont un lien etroit avec les ee math matiques, comme les soci t s de statistique, les soci t s biom triques, e ee ee e les soci t s de math matiques industrielles ou encore, aux U. S. A., la Society ee e for Quality Control qui compte d j` six mille membres en 1953. ea Dans ce contexte Kurepa sinterroge sur la d nition du math maticien et, e e consid rant l volution des emplois des gens soccupant de math matiques , e e e il consid` re ce quil appelle des math maticiens au sens large dont font e e partie les math maticiens pratiques que sont les ing nieurs . Cela le conduit e e ` ` a questionner, sinon a rejeter, tout clivage entre math matiques pures et math e e matiques appliqu es. e La r daction m me du questionnaire diffus par la CIEM traduit un axe e e e ` fort de cette enqu te, a savoir la relation nouvelle que les math matiques e e entretiennent avec les autres disciplines. Il est int ressant, tout dabord, de e ` relever le vocabulaire employ dans les rapports a ce propos ; ils parlent e d enchev trement , d interconnexions des math matiques aux autres discie e plines. Ces autres disciplines appartiennent tout autant aux sciences dites dures quaux sciences humaines ou aux sciences economiques. Les rapports en multiplient les exemples qui, en particulier aux U. S. A., apparaissent egalement marqu s par les domaines issus de recherches men es durant la guerre. Ils e e notent, de plus, que cet appel de math matiques dans tant de disciplines nest e pas la cons quence dune propagande ou dune campagne de publicit des e e ` math maticiens, mais correspond a une v ritable demande autonome de ces e e diff rents champs. e ` Kurepa et les autres rapporteurs insistent alors sur le fait que cest a partir de ces mutations de lactivit math matique, des evolutions du contenu m me e e e de la notion de math matiques et de math maticien, quil sagit de penser e e lenseignement des math matiques. Les applications, les besoins dun modern e consumer of mathematics 5 ), impliquent la n cessit dune r forme de la e e e formation int grant de nouveaux contenus th oriques fondamentaux ; ce dont e e il y a besoin cest dapplications des math matiques modernes ou de math e e matiques appliqu es modernes. Sont cit es, par exemple, lid e de d pendance e e e e stochastique, lintroduction de lal atoire, la recherche op rationnelle, lanalyse e e ` ` num rique a cause de nouvelles machines a calculer, la statistique, la th orie e e de linformation Mais il y a plus ; il est n cessaire, afrment-ils, de d passer lid e dominante e e e selon laquelle lenseignement secondaire devrait se r duire aux math matiques e e pures et au seul raisonnement purement d ductif. Ce dernier, en effet, ne saurait e
5
) Cette expression est employ e par van Dantzig dans son rapport. e
257
etre transf r dans les diff rents domaines des math matiques appliqu es et ee e e e on laisse alors de c t , par exemple, la logique inductive ou la notion de oe plausibilit . e Au-del` de ces constats sur la place des math matiques et des math a e e maticiens dans la soci t , sur les cons quences attendues et n cessaires ee e e sur lenseignement des math matiques, un seul rapport rend compte de e changements effectifs dans lenseignement secondaire. Il sagit du rapport ` e de la sous-commission des Pays-Bas, seul pays ou, par exemple, des el ments e de statistique ont et introduits dans lenseignement secondaire ; pour dautres pays, les exemples de changement qui sont avanc s dans les rapports concernent e lenseignement sup rieur, comme en France avec lInstitut de statistique de e Paris. Le rapport de Weinberger mentionne, par contre, les efforts, a priori non couronn s de succ` s, des congr` s internationaux de statistique qui se sont e e e tenus depuis la guerre pour lintroduction dun enseignement de statistique dans les lyc es. e STOCKHOLM, 1962 : WHICH
SUBJECTS IN MODERN MATHEMATICS AND WHICH
APPLICATIONS IN MODERN MATHEMATICS CAN FIND A PLACE IN PROGRAMS OF SECONDARY SCHOOL INSTRUCTION ? 6 )
Cette deuxi` me enqu te, lanc e en 1958 par la CIEM, se positionne e e e ` diff remment par rapport a notre sujet. Son titre m me d tache les applications e e e de ce qui serait le corpus de math matiques modernes. Il y aurait dune part e les sujets des math matiques, et dautre part les applications de ces sujets. e ` e Cette mise a l cart, qui tranche par rapport au ton de lenqu te pr c dente, e e e est conrm e par le contenu de lenqu te. e e Lenqu te sappuie sur 21 rapports nationaux dont seuls les noms des rape porteurs, avec leurs pays, gurent dans la revue 7 ). Le rapport de synth` se e pr sent par J. G. Kemeny [1964] au Congr` s international de 1962 souligne le e e e ` degr de similarit de ces contributions nationales a la fois quant aux proposie e tions pour introduire de nouveaux sujets math matiques dans lenseignement et e
6 ) Cette enqu te de la CIEM est annonc e avec un titre en francais dans la lettre circulaire e e du 28 juin 1958 du bureau de la CIEM aux dirigeants des sous-commissions nationales, publi e e dans le tome 5 (1959) de LEnseignement Math matique. Le rapport de Stockholm sur lenqu te e e est paru dans le tome 10 (1964) de la revue [Kemeny 1964]. 7 ) Les pays qui ont particip a lenqu te sont lAllemagne, lAngleterre, lArgentine, e ` e lAustralie, le Danemark, la Finlande, la France, la Gr` ce, la Hongrie, lInde, Isra l, lItalie, e e le Luxembourg, la Norv` ge, les Pays-Bas, la Pologne, le Portugal, la Sierra Leone, la Su` de, la e e Suisse et les U. S. A.
258
H. GISPERT
pour le peu dattention accord a lenseignement des applications des math e ` e matiques 8 ). Dans tous ces rapports, il nest en effet envisag quun seul sujet appliqu , e e la statistique. Ils proposent lintroduction des probabilit s et de la statistique e qui apparaissent toujours coupl es, les probabilit s etant consid r es comme e e ee sujet de math matiques pures et la statistique comme sujet de math matiques e e appliqu es. On est loin de la diversit des r exions de 1954. e e e Certains des rapporteurs nationaux s meuvent de la situation quils ont e e pu constater dans leur pays o` , alors quun enorme effort a et fait pour u am liorer lenseignement des math matiques pures, le sujet des math matiques e e e e appliqu es a apparemment et oubli . Les Scandinaves ont, semble-t-il, une e e place davant-garde dans ce souci de promotion des applications : leurs rapports indiquent en effet que le Scandinavian Committee for the Modernizing of School Mathematics a adopt en 1960 un projet en cinq points dont le premier e ` est ici remarquable. Il sagit d tudier les besoins en math matiques a la fois e e ` pour le monde industriel et pour luniversit et de proposer a partir de cela e de nouveaux curricula. Dans son rapport de synth` se, Kemeny souligne les insufsances not es e e dans les diff rents rapports nationaux, relaie les inqui tudes des rapporteurs e e et pr sente des recommandations particuli` res. Il propose que lenseignement e e des applications des math matiques soit une des priorit s dans les etudes de e e ` la CIEM pour les quatre ann es a venir ; il sugg` re a cette n, mis a part e ` e ` ` la statistique, trois types dapplications : les applications des math matiques a e la physique, la programmation lin aire peut- tre le seul exemple, ecrit-il, e e o` l tudiant pourrait constater un lien v ritable avec les sciences sociales u e e ` et le libre usage des machines a calculer dans lenseignement. Reprenant des points de lenqu te pr c dente de 1954, Kemeny insiste nalement sur le e e e fait que lenseignement de math matiques appliqu es suppose de d velopper e e e dans lenseignement de nouvelles habitudes de pens e qui diff` rent souvent e e ` de celles a luvre dans les math matiques abstraites. e Le rapport se termine enn par la recommandation de lancer plusieurs etudes, dont une sur un probl` me notoirement n glig dans le pass et qui e e e e concerne les math matiques appliqu es : How can the teaching of applied e e mathematics in our high schools be modernized ?
8 ) Lexpression adopt e par Kemeny au d but de la quatri` me partie de son rapport consacr e e e e e aux applications des math matiques est particuli` rement forte : It is painfully clear, in reading e e the 21 national reports, that [Kemeny 1964, 166].
259
UNE DIFFERENCE DE TON MANIFESTE : EVOLUTION DE LA CIEM OU DIFFICULTES OBJECTIVES ? Les applications des math matiques et leur enseignement nont ainsi ni e la m me place ni le m me r le dans ces deux etudes de la CIEM. Cette e e o ` diff rence de ton entre l tude de 1954 et celle de 1962 est-elle due a une e e evolution dans les positions ou les priorit s de la CIEM ? Ou bien est-elle e ` ` due a des difcult s objectives a la prise en compte de ces applications dans e les milieux math matiques traditionnels responsables de lenseignement ? e Il faut noter tout dabord que lenqu te dAmsterdam de 1954 aurait du e 9 etre poursuivie et elargie ). Ce ne fut pas le cas, comme en t moigne le e compte rendu dune r union du comit ex cutif de la CIEM en 1955 paru e e e dans LEnseignement Math matique [CIEM 1955, 198201]. Elle fut mise de e c t , puis abandonn e, a la suite dautres propositions denqu tes de Hans oe e ` e Freudenthal sur des points beaucoup plus techniques ou didactiques et sur des sujets math matiques plus cibl s dont la g om trie. e e e e En m me temps, dans cette etude de 1954, le rapport de la souse commission des U. S. A. soulignait les limites de la nouvelle situation cr ee e par le d veloppement dapr` s-guerre des applications des math matiques. Le e e e nouvel int r t pour les applications les plus modernes des math matiques dans ee e un si grand nombre de disciplines, limplantation de laboratoires de math e matiques dans lindustrie, l conomie, etc., navaient pas encore provoqu de e e changement majeur dans le milieu math matique am ricain, la plupart des e e math maticiens ayant un poste dans le monde universitaire. De m me, Jeremy e e Kilpatrick mentionne que les r formateurs, aux U. S. A., etaient majoritairement e des math maticiens universitaires impliqu s plut t dans des recherches de e e o math matiques pures que dans des recherches de math matiques appliqu es e e e [Kilpatrick 1996]. ` Cependant, malgr labandon a mon avis signicatif de lenqu te e e de 1954, la CIEM ne va pas d serter le champ des applications des math e e ` Belgrade un colloque matiques. En premier lieu, en 1960, elle organise a sur la coordination des enseignements de math matiques et de physique et e engage une enqu te sur lenseignement des math matiques pour les physiciens. e e Cette enqu te est pr sent e par Charles Pisot au Congr` s international des e e e e ` math maticiens a Moscou en 1966. Dans son rapport, publi par LEnseie e gnement Math matique, il souligne que la pr occupation fondamentale de e e
9 ) Voir le compte rendu dactivit de la CIEM apr` s le congr` s dAmsterdam : Le rapport e e e de la seconde enqu te tentait de dresser un inventaire de lapplication des math matiques aux e e activit s et m tiers les plus vari s. Lenqu te, qui nen est qu` ses d buts, sera poursuivie et un e e e e a e rapport sera pr sent au prochain congr` s, a Edimbourg. [Behnke 195154, 90] e e e `
260
H. GISPERT
lenseignement au niveau 1618 ans doit etre la coordination aussi etroite que possible entre les math maticiens et la physique. L` , ecrit-il, est le nud de e a lenseignement secondaire [Pisot 1966, 203]. ` Il semble ici que la CIEM privil gie le lien a la physique, le lien avec les e physiciens. Or, les recommandations que Andr Lichn rowicz, alors pr sident e e e de la CIEM, avance dans son rapport sur la p riode 19631966 insistent e ` pourtant sur limportance quil y aurait a prendre en compte les probl` mes e ` propres de liaison qui se posent a lenseignement des math matiques non e seulement pour les sciences de la nature et pour la technique, mais aussi pour les sciences economiques et sociales [Lichn rowicz 1966]. Mais peut- tre e e ` y a-t-il une r elle difcult a d passer le stade des intentions et a trouver e e ` e au sein de la CIEM, ou des diff rents milieux math matiques nationaux, e e ` des hommes pour mener a bien de telles initiatives. En effet, en 1967, la ` CIEM cherche a lancer en collaboration avec lUNESCO un projet douvrage ` qui illustre lapplication des math matiques a diverses sciences, mais il y a e quelque difcult a trouver un math maticien qui se charge de sa r alisation. e ` e e De plus, toujours en 1967, des propositions faites par la Grande-Bretagne pour ` d ventuelles etudes de la CIEM sur les applications des math matiques a la e e biologie et aux recherches sociales, ou sur limportance du calcul automatique ` a tous les niveaux scolaires, ne sont pas retenues [CIEM 1967]. Cest sur un tout autre plan quest d battue, au colloque dUtrecht organis e e par la CIEM en 1967, la question de lutilit des math matiques pour les e e ` autres sciences et des cons quences a en tirer pour son enseignement. Son e titre m me, How to teach mathematics so as to be useful , indique un e glissement dint r t ; les pr occupations de ce colloque sont essentiellement ee e dordre didactique et du registre de la classe, et les r f rences aux contextes ee social, economique et m me math matique not es pr c demment semblent e e e e e ` avoir disparu. Ce colloque ne correspond a aucune enqu te pr alable de e e la CIEM et, il me semble important de le souligner, nest ni annonc , ni e comment dans LEnseignement Math matique. Cest dans le premier num ro e e e de la toute nouvelle revue Educational Studies in Mathematics fond e par e Hans Freudenthal, alors pr sident de la CIEM, que sont publi s les Actes de e e ce colloque. Il semble que nous ayons l` un tournant 10 ) dans la nature des a liens que la CIEM a entretenus avec la revue LEnseignement Math matique. e ` Dans son intervention a ce colloque, Henry Pollak (des Bell Telephone Laboratories), un des acteurs les plus pr sents des initiatives de ces ann es e e
10 ) Daniel Coray fait remarquer que 1967 est egalement lann e du d c` s de Jovan Karamata, e e e qui dirigeait la revue depuis le d but de la 2e s rie et qui avait aussi cr e la collection des e e e Monographies de lEnseignement Math matique. e
261
sur lenseignement des math matiques, souligne le nouveau probl` me auquel e e ` ont a faire face les educateurs : consid rant qu` la n des ann es 1960 it e a e is just not possible to object to applications of mathematics as a part of the educational process. Why therefore do we have so much difculty ? Why was it necessary to hold this meeting ? [Pollak 1968] Une premi` re difcult e e ` tient a la contradiction suivante que Hans Freudenthal met en avant dans son intervention douverture. En effet, indique-t-il, apr` s avoir pr cis quenseigner e e e des math matiques de mani` re a ce quelles soient utiles nest pas la m me e e ` e chose quenseigner des math matiques utiles : e
Useful mathematics may prove useful as long as the context does not change, and not a bit longer, and this is just the contrary of what true mathematics should be. Indeed it is the marvellous power of mathematics to eliminate the context. [] In an objective sense the most abstract mathematics is without doubt the most exible. In an objective sense, but not subjectively []. 11 ) [Freudenthal 1968, 5]
` Do` la r ponse que Willy Servais apporte a la question-titre du colloque, u e ` qui montre toute la diversit des biais possibles dans les r f rences a cette e ee notion dutilit pour ce qui est des priorit s a d gager pour les curricula : e e ` e assurer les conditions dapprentissage qui font pratiquer la math matique e pour ce quelle est : une activit cr atrice de structures qui permettent de e e saisir la r alit [Servais 1968, 53]. e e Le colloque met en avant un deuxi` me type de difcult , cette fois e e e dordre p dagogique ou didactique, qui concerne tant les el` ves ou les e ` etudiants que les enseignants face a ce que lon ne nomme pas encore ` mod lisation mais math matisation moderne. Face a la port e et au style e e e des applications contemporaines des math matiques et des proc d s de cette e e e math matisation moderne, le professeur ressent un sentiment d sagr able e e e ` dincertitude, dincomp tence, de dilettantisme et cherche a eviter ces situations e ` a la limite. De plus, obstacle suppl mentaire, il nest pas certain comme e le fait remarquer A. Z. Krygovska [1968] quil soit ais de trouver des e types de situation que lon puisse math matiser avec prot en classe. Enn, e une troisi` me difcult d battue lors du colloque tient cette fois aux seuls e e e etudiants dont lattitude dans les classes montre quils nattendent de leurs enseignants de math matiques que des proc d s qui leur donnent rapidement e e e acc` s aux r sultats. e e
11 ) Les math matiques utiles peuvent se montrer utiles tant que le contexte ne change pas, e et pas un moment de plus, ce qui est juste le contraire de ce que les vraies math matiques e devraient etre. En effet, nest-ce pas le merveilleux pouvoir des math matiques que d liminer le e e contexte ? [] En un sens objectif les math matiques les plus abstraites sont indubitablement les e plus exibles. En un sens objectif, mais pas subjectivement [].
262
H. GISPERT
Quelles ques soient les difcult s p dagogiques, il semble n cessaire e e e ` darriver a les surmonter. Cest ce dont t moigne la conclusion de lintervention e d j` mentionn e de Freudenthal intitul e Why to teach mathematics so as ea e e to be useful ? . Reprenant des accents d j` entendus soixante ans plus t t ea o dans des conf rences reproduites dans LEnseignement Math matique, Hans e e Freudenthal sinqui` te de ce que les utilisateurs de math matiques pourraient e e dessaisir les math maticiens de lenseignement de leur discipline si ceuxe ` ci narrivaient pas a enseigner les math matiques de sorte quelles soient e utilisables. Ce serait alors la n de toute education math matique. e Cela dit, un an plus tard, au premier Congr` s international sur lenseignee ` ment des math matiques (ICME) quorganise la CIEM a Lyon, cette question e ne fait pas recette. Seuls deux intervenants, sur la quarantaine qui prirent ` la parole, trait` rent dapplications ou de math matisation. Lun deux etait a e e nouveau Pollak qui analysait dans son intervention [Pollak 1969] un certain ` nombre de probl` mes renvoyant a des situations de math matisation. e e
2. LES
COLLOQUES DE LOECE ET DE LOCDE :
LE
RECOURS AUX
APPLICATIONS DANS LENSEIGNEMENT DES MATHEMATIQUES 12 ) Dans son rapport dactivit pour la p riode 19591962, la direction e e sortante de la CIEM insiste sur la participation de plusieurs de ses membres aux initiatives de lOECE, puis 13 ) de lOCDE, pour la modernisation de lenseignement math matique [ICMI 1963]. Ces initiatives commencent avec e le colloque de Royaumont en 1959 en France, puis celui de Dubrovnik et celui dAth` nes en 1963. Les membres de la CIEM engag s dans ces colloques e e sont des responsables importants de la commission. Marshall H. Stone (vice` pr sident puis pr sident de la CIEM de 1955 a 1962) est lorganisateur du e e ` colloque de Royaumont ; quant a Ath` nes, Andr Revuz (secr taire de la e e e ` CIEM de 1959 a 1962) est responsable du texte francais de louvrage issu du e colloque edit par Howard F. Fehr, directeur du Department of Mathematical Education de lUniversit de Columbia, impliqu dans les activit s de la CIEM e e e et consultant pour lOCDE.
12 13
) Il sagit du titre dune des sessions de la conf rence dAth` nes de lOCDE en 1963. e e
) LOrganisation de coop ration et de d veloppement economiques a pris en 1961 la rel` ve e e e e e de lOrganisation europ enne de coop ration economique (qui avait et cr ee au lendemain de e e la seconde guerre mondiale pour administrer laide des Etats-Unis et du Canada dans le cadre du Plan Marshall de reconstruction de lEurope).
263
Limplication dun organisme tel que lOCDE dans un mouvement de modernisation de lenseignement des math matiques, en particulier dans celui e qui aboutit aux diff rentes r formes dites des math matiques modernes dans le e e e monde, pourrait avoir de quoi surprendre. Le colloque dAth` nes sen explique, e une intervention soulignant que la r forme de lenseignement sest faite sous e limpulsion de trois forces diff rentes : celle des math maticiens qui voulaient e e une pr sentation des math matiques en accord avec les th ories modernes, celle e e e des psychologues de ladolescence qui ont montr la possibilit dun meilleur e e apprentissage, et celle des dirigeants de lindustrie nous retrouvons ici lOCDE qui demandaient une meilleure pr sentation des applications des e math matiques 14 ). e Jexaminerai ici plus particuli` rement ce colloque dAth` nes et louvrage e e [OCDE 1963] qui en est issu. Quatri` me ouvrage de la direction des affaires e scientiques de lOCDE dans la s rie Pour un enseignement r nov des e e e sciences , consacr a lenseignement moderne des math matiques , ce livre e` e dont le titre traduit lambition : Math matiques modernes ; guide pour e enseignants contient en effet un chapitre consacr aux applications. La e pr sence dun tel chapitre, parmi les six du livre, intitul Applications dans e e la modernisation des math matiques, rev t un int r t tout particulier. La e e ee prise en compte des applications est en fait afrm e plusieurs fois dans e louvrage, et pas seulement dans le chapitre qui y est sp ciquement d volu. e e Le chapitre 5, o` sont expos es les conclusions et les recommandations du u e colloque, y consacre, par exemple, plusieurs items dont je donne ici les id es e directrices. Il sagit tout dabord dinclure dans les programmes les th ories et mati` res e e qui ont acquis une grande importance, soit par leur role unicateur dans lexpos des math matiques, soit par lint r t particulier quils pr sentent pour e e ee e les applications dans dautres sciences, en particulier les sciences physiques. Une liste est donn e dans laquelle les deux registres envisag s ne sont pas e e s par s ; citons : les ensembles, les relations et les fonctions, les principes et e e les structures alg briques fondamentales, les espaces vectoriels et lalg` bre e e lin aire, le calcul diff rentiel, le calcul des probabilit s et la statistique, les e e e th ories math matiques qui ont un lien avec les calculatrices electroniques. e e On y souligne limportance, dans les programmes, de traiter dans une seule et m me etude les structures logiques et les probl` mes. e e
14 ) Voir dans [OCDE 1963, 194] le r sum de la discussion qui a suivi lintervention de e e H. Pollak.
264
H. GISPERT
En second lieu, il faut am liorer les m thodes denseignement et d velopper e e e e ` laptitude des el` ves a rechercher la solution de probl` mes concrets dans un e cadre math matique. Ce dont il est question, en fait, cest de d velopper une e e approche de la mod lisation math matique, appel e ici chercher le mod` le , e e e e ` le mot mod` le etant dailleurs a cette occasion utilis pour la premi` re fois, e e e du moins dans la litt rature sur lenseignement des math matiques que jai e e pu consulter. Dans cette perspective la conf rence invite lensemble des pays e ` a constituer des s ries syst matiques dexemples dapplications des math e e e matiques adapt es a lenseignement secondaire, ce qui suppose une recherche e ` permanente dans laquelle il est urgent de sengager. Enn, la r solution nale insiste enorm ment sur limportance des ape e plications des math matiques modernes dans la formation des professeurs, e sur la valeur de lenseignement des math matiques par les applications, sur e la n cessit de justier les notions math matiques par leurs applications et e e e ` de faire comprendre que les math matiques sont utiles a la soci t . Quant e ee ` au champ des applications, limportance des liens avec les physiciens est a nouveau mise en avant. ` Cette insistance sur les applications, sur leur n cessaire int gration a lenseie e gnement math matique, tant dans son esprit et ses contenus que dans ses e m thodes, est dautant plus remarquable quelle apparat comme collectie ` vement port e par le colloque. Et cela, dans un colloque ou parmi les orateurs e de premi` re importance gurent certes Henry Pollak omnipr sent, nous e e ` lavons dit, a tous ces congr` s ou conf rences , mais aussi Andr Revuz, e e e Willy Servais, secr taire de la Commission internationale pour l tude et e e lam lioration de lenseignement math matique (CIEAEM), cr ee apr` s-guerre, e e e e ou Georges Papy, toutes personnalit s tr` s engag es et tr` s repr sentatives du e e e e e mouvement de r forme des math matiques modernes, en France et en Belgique e e notamment. Cependant, cette insistance dans les conclusions et recommandations de la conf rence semble quelque peu volontariste. Elle ne re` te pas, en particulier, e e la part que les intervenants, tant dans les conf rences que dans les discussions e qui suivirent, ont r ellement r serv e a ces questions dapplications des math e e e ` e matiques 15 ).
15 e ) Ainsi, si une journ e a et consacr e a Ath` nes au th` me recours aux applications dans e e ` e e lenseignement des math matiques , le programme de cette journ e apparat bien mince et ces e e questions ne resurgissent que pour les conclusions.
265
3. ET
QUEN A-T-IL ETE ALORS DE LIMPACT DE
BOURBAKI
SUR LES REFORMES DES MATHEMATIQUES MODERNES ? Le nom commun le plus souvent associ aux math matiques dites modernes e e est celui de structures, le nom propre, en France mais aussi dans le monde, est celui de Nicolas Bourbaki qui lui est si etroitement li . Or, Bourbaki ne sest e pas int ress aux math matiques appliqu es et a ignor des th ories math e e e e e e e ` matiques si essentielles et si pr sentes dans les applications, a commencer par e les probabilit s et la statistique. Ny a-t-il pas alors une contradiction entre e ce souci appuy des applications des math matiques et les images qui collent e e au mouvement de r forme des math matiques modernes ? e e Il faut noter, tout dabord, que malgr les conditions qui ont provoqu sa e e naissance, Bourbaki, comme groupe de math maticiens, ne sest jamais vraie ment int ress aux questions denseignement, de curriculum ou de p dagogie, e e e ni dans le sup rieur, ni encore moins dans le secondaire. De plus, il serait e faux, y compris en France et en Belgique, de limiter la modernisation de ` lenseignement des math matiques a la prise en compte de la nouvelle archie tecture des math matiques introduites par la notion de structures. e Cette modernisation est par exemple d nie en trois points dans le e guide pour les enseignants [OCDE 1963]. Si le premier point est bien : unier lensemble du sujet en insistant sur les structures fondamentales , ` les deux autres points rel` vent des applications et consistent a inclure dans e les programmes les th ories qui ont acquis une grande importance du fait e de leurs applications dans dautres sciences (entre autres la statistique) et ` ` a am liorer les m thodes denseignement par le recours a la r solution de e e e probl` mes concrets dans un cadre math matique. e e Il ne semble pas quil y ait eu, entre les diff rents acteurs de la promotion e des math matiques modernes dans lenseignement, de d saccords fondamene e taux sur cette n cessaire prise en compte des applications. Tous en soulignent e limportance, soit dans des prises de position individuelles, soit en sassociant ` a des manifestations qui insistent tout particuli` rement sur ce point. M me e e 16 la CIEAEM, qui dans ses propres publications ) ne semble pas concern e e par cette dimension, sy trouve de fait associ e par la pr sence et le r le e e o de son pr sident Servais au sein de la CIEM et au colloque de lOCDE e de 1963. Il y a par contre d bat sur la facon de d cliner cette ouverture aux e e applications. Quelles math matiques sont utiles pour quelles applications ? e
16
) Voir par exemple [Piaget & al. 1955].
266
H. GISPERT
Enseigner les math matiques an quelles soient utiles et applicables peut e signier prioritairement et quelquefois uniquement, dans la r alisation e ` effective des principes apprendre a abstraire et non viser les applications ` directes. Cette conception conduit, de fait, a faire limpasse sur lintroduction de nouvelles th ories (statistique par exemple), sur lapprentissage de la e ` mod lisation par le traitement de cas concrets, sur linitiation a dautres formes e de raisonnement que la stricte logique d ductive, etc. e La facon de concevoir lutilit et lapplicabilit des math matiques, et e e e donc leur enseignement, d pend en fait des champs dapplications vis s, des e e disciplines concern es. Si la CIEM et lOCDE insistent dans les ann es 1960 e e sur la physique, il nen est pas de m me, par exemple, de la Commission e minist rielle francaise de r forme de lenseignement des math matiques e e e dirig e par Andr Lichn rowicz. Dans sa premi` re d claration ofcielle, la e e e e e Commission relativise les liens de tout temps privil gi s entre les math e e e matiques et la physique et insiste, au contraire, sur la diversication des disciplines et des champs de lactivit humaine qui ont besoin des math e e matiques.
La math matique joue un r le privil gi pour lintelligence de ce que nous e o e e nommons le r el, le r el physique comme le r el social. Notre math matique e e e e s cr` te par nature l conomie de pens e et, par l` , permet seule de classer, de e e e e a e dominer, de synth tiser []. La math matique a et , depuis toujours, discipline e e auxiliaire des sciences physiques et de lart de ling nieur. Elle est devenue e d sormais, au m me titre, discipline auxiliaire, aussi bien dune grande partie des e e sciences biologiques et m dicales que de l conomie et des sciences humaines. e e [] Elle porte partout t moignage du fonctionnement m me de notre esprit. e e [Commission minist rielle 1967] e
De l` limportance donn e a luniversalit du langage math matique et des a e ` e e structures, qui rendent les math matiques contemporaines inniment plus e applicables , au d triment des attributs plus traditionnels des math matiques e e dites classiques que sont le calcul innit simal et la g om trie. Le choix e e e ` des th ories, des outils a privil gier dans lenseignement des math matiques e e e est en effet li aux questions pr c dentes. Mais ce choix fait d bat, y compris e e e e lorsquon ne consid` re quun seul champ dapplication. e ` e Un bon exemple en est le lien a la physique qui a et en France, durant le mouvement de r forme des math matiques modernes, un point tr` s litigieux. e e e Les scientiques francais, par le biais de leurs associations de sp cialistes, e protest` rent contre les nouveaux programmes de math matiques des sections e e scientiques des lyc es issus de la r forme des math matiques modernes. Ils e e e ` sinqui taient du nouveau visage de lenseignement scientique en France a e la suite des effets de ce quils appellent en 1970 la r novation p dagogique et e e
267
quils caract risaient comme une reconversion aux math matiques modernes e e et une r duction accrue de la place accord e aux sciences exp rimentales. e e e Ils d nonc` rent lenvahissement par les math matiques d lib r ment les plus e e e e ee abstraites ainsi que
cette ecole de dogmatisme [qui] a pour dernier souci, et de motiver ses ` abstractions par r f rence initiale a quelque probl` me concret, et de veiller ee e ` a fournir aux autres disciplines les outils math matiques (ou, si lon pr f` re, e ee de calcul) qui leur sont n cessaires. 17 ) e
Ces d bats et ces tensions provoqu` rent une rupture dans la commission e e Lichn rowicz. Charles Pisot, le rapporteur de lenqu te de 1966 de la CIEM e e sur la formation des ing nieurs, d missionna de la Commission (dont il etait e e ` membre) et contribua a la cr ation dune nouvelle association, lUnion des e professeurs et utilisateurs de math matiques (UPUM) qui milita pour pr server e e une formation math matique traditionnelle des scientiques et ing nieurs. e e ` En fait, malgr tous les discours et principes favorables aux applications, a e des applications modernes des math matiques dites modernes, discours tenus, e ` nous lavons vu, au colloque dUtrecht de la CIEM, a la r union dAth` nes, e e dans la Commission Lichn rowicz, le mouvement des math matiques modernes e e se solde effectivement, dans la r alit des programmes, par un affaiblissement e e des anciennes solidarit s disciplinaires. Et cela, il est important de le souligner, e e e sans quelles aient et r ellement remplac es par de nouvelles. e Au-del` de ce constat, quelles conclusions peut-on tirer de l tude de cette a e p riode ? Tout dabord, Geoffroy Howson la dit, il faut savoir lire les textes e du pass m me dun pass r cent pour saisir les enjeux dune p riode. e e e e e e ` Les r formateurs des ann es 19501960 ont et face a un enjeu nouveau : la e e scolarisation secondaire de tous. Au d but du vingti` me si` cle, en France par exemple, les seules solidarit s e e e e disciplinaires pour la grande masse des enfants, qui ne connaissent alors que e lenseignement primaire el mentaire 18 ), sont larpentage et le dessin lin aire. e Pour les autres, ceux qui poursuivent leur scolarit au-del` de 12 ou 13 ans, e a sil y a effectivement dans lenseignement math matique secondaire un lien e ` ` revendiqu a la physique, a la m canique, a lastronomie, cela ne concerne que e` e
17 ) Communiqu de 1970 de lUnion des physiciens, de la Soci t francaise de physique et e ee de la Soci t chimique de France, publi dans [Hulin 1992, 41]. ee e 18 ) La situation francaise est dune certaine facon particuli` re; les Etats-Unis, par exemple, e connaissent d` s le d but du si` cle une ecole unique, gratuite et obligatoire jusqu` 14 ans. Mais e e e a la diff rence de registre dans les solidarit s disciplinaires de lenseignement des math matiques, e e e ` que jexplicite dans le cas francais, demeure a mon avis pertinente pour tous les pays. Tous ` ` connaissent en effet, a un niveau scolaire ou un autre, une s paration entre des enseignements a e nalit s avant tout pratiques et dautres dont la nalit est plus lib rale; voir [Nabonnand 2003]. e e e
268
H. GISPERT
` les classes de second cycle de lenseignement des lyc es (qui a cette epoque e etait payant) ; dans lenseignement primaire sup rieur, o` le calcul innit simal e u e nest pas d velopp , ce sont les liens avec le monde du commerce, de la e e banque, de la m canique des machines et non pas de la m canique rationnelle e e qui sont afrm s. Si lon poursuit au-del` de la premi` re guerre mondiale e a e l tude du cas francais, on constate que l volution de lenseignement math e e e matique secondaire dans lentre-deux-guerres prend de plus en plus de distance avec les applications revendiqu es au d but du si` cle. La vocation lib rale et e e e e d sint ress e de cet enseignement destin a une elite na pas fait bon m nage e e e e ` e avec les ambitions des r formateurs du d but du si` cle. e e e Au lendemain de la seconde guerre mondiale, la question des applications dans lenseignement math matique se pose, pour tous les pays, avec une e ` ` nouvelle dimension. Comme le souligne M. H. Stone en 1952 a Rome, a loccasion de la reprise des activit s de la CIEM comme commission de la e toute nouvelle Union math matique internationale (UMI), cette question se e pose dans le cadre dune instruction populaire obligatoire que vont devoir assumer tous les pays [Stone 1953]. Les d bats sur les applications et les e mises en uvre eventuelles des projets en deviennent singuli` rement plus e complexes. ` Face a ces d s, quels pouvaient etre les appuis des r formateurs, de la e e CIEM ou dailleurs, conscients de la n cessit du d veloppement des math e e e e matiques comme discipline de service et de lintroduction de cette dimension dans lenseignement ? Ils ne viendront pas du monde math matique lui-m me. e e Comme lavait laiss penser le rapport de la sous-commission des U. S. A. e pr sent par Kemeny, les changements dans le monde math matique apr` s e e e e 19 guerre ne furent pas aussi radicaux ) que le laissait supposer la premi` re e enqu te de la CIEM. Tant du point de vue des int r ts math matiques que des e ee e d bouch s professionnels, les math matiques acad miques dominent toujours e e e e le monde math matique et plus particuli` rement la partie int ress e par les e e e e questions denseignement. Les appuis ne sont pas venus non plus du monde enseignant qui sest av r tr` s conservateur. Les difcult s li es a la formation ee e e e ` des matres dans le domaine des applications des math matiques sont dailleurs e sans cesse soulign es dans les divers colloques ou enqu tes. e e Ce constat appelle alors plusieurs questions qui ont d j` et pos es dans ea e e ce Symposium. Quel biais la CIEM apporte-t-elle dans l tude que lon peut e faire de l volution et de lhistoire de lenseignement des math matiques dans e e
19 ` ) On peut lire a ce sujet, dans larticle dAmy Dahan [1996] d j` cit , la partie quelle ea e ` ` consacre a la Conference on training in applied mathematics tenue a luniversit de Columbia e (New York) en 1953 (voir en particulier p. 196).
269
le vingti` me si` cle ? De quoi, de qui la CIEM est-elle repr sentative ? Voil` , e e e a ` me semble-t-il, des questions tout a fait pertinentes quil serait int ressant de e ` traiter dans la perspective du centenaire a venir de la CIEM.
BIBLIOGRAPHIE BEHNKE, H. Lactivit de la Commission internationale de lEnseignement math e e matique. LEnseign. Math. 40 (195154), 9091. BOURLET, C. La p n tration r ciproque des math matiques pures et des math matiques e e e e e appliqu es dans lenseignement secondaire. LEnseign. Math. 12 (1910), 372387. e ` CIEM. Compte rendu de la r union du Comit ex cutif de la CIEM a Gen` ve, le e e e e 2 juillet 1955. LEnseign. Math. (2) 1 (1955), 196202. ` Compte rendu de la s ance de la CIEM tenue a Utrecht, le 26 ao t 1967. e u LEnseign. Math. (2) 13 (1967), 243246. COMMISSION MINISTERIELLE FRANCAISE DE REFORME DE LENSEIGNEMENT DES MATHE MATIQUES (sous la dir. de A. Lichn rowicz). Rapport pr liminaire. Bulletin de e e lAssociation des professeurs de math matiques 258 (1967), 246248. e DAHAN, A. Lessor des math matiques appliqu es aux Etats-Unis : limpact de la deue e xi` me guerre mondiale. Revue dhistoire des math matiques 2 (1996), 149213. e e FREUDENTHAL, H. Why to teach mathematics so as to be useful ? Educational Studies in Mathematics 1 (1968), 38. HULIN, M. Le mirage et la n cessit . Presses de lEcole normale sup rieure, 1992. e e e ICMI. Report for the period 195962. LEnseign. Math. (2) 9 (1963), 105114. KEMENY, J. G. Which subjects in modern mathematics and which applications in modern mathematics can nd a place in programs of secondary school instruction ? LEnseign. Math. (2) 10 (1964), 152176. KILPATRICK, J. R former les programmes de math matiques aux U. S. A. depuis 1900 : e e r alit et imaginaire. In : Belhoste, Gispert, Hulin ( ds), Les sciences au lyc e, e e e e 249258. VuibertINRP, Paris, 1996. KRYGOVSKA, A. Z. Processus de la math matisation dans lenseignement. Educational e Studies in Mathematics 1 (1968), 916. ` e KUREPA, G. Le r le des math matiques et du math maticien a l poque contemporaine. o e e Rapport g n ral. LEnseign. Math. (2) 1 (1955), 93111. e e LICHNEROWICZ, A. Commission internationale de lenseignement math matique. Rape port sur la p riode 19631966. LEnseign. Math. (2) 12 (1966), 131138. e NABONNAND, P. [2003] Les d bats autour des applications des math matiques dans e e les r formes de lenseignement secondaire au d but du vingti` me si` cle. This e e e e volume, 229249. OCDE. Math matiques modernes. Guide pour enseignants. OCDE, 1963. e PIAGET, J., E. W. BETH, J. DIEUDONNE, A. LICHNEROWICZ, G. CHOQUET et C. GATTEGNO. Lenseignement des math matiques. Delachaux & Niestl , Neuch tel, e e a 1955. PISOT, C. Rapport sur lenseignement des math matiques pour les physiciens. LEnseign. e Math. (2) 12 (1966), 201216.
270
H. GISPERT
POLLAK, H. On some of the problems of teaching applications of mathematics. Educational Studies in Mathematics 1 (1968), 2430. How can we teach applications of modern mathematics ? Educational Studies in Mathematics 2 (1969), 393403. SERVAIS, W. Comment enseigner la math matique pour quelle soit utile ? Probl matique e e et axiomatique. Educational Studies in Mathematics 1 (1968), 3753. STONE, M. H. [1953] LUnion math matique internationale et ses activit s. Rapport e e sur la 1re assembl e g n rale (Rome, 68 mars 1952). LEnseign. Math. 39 e e e (19421950), 156161.
APPLICATIONS OF MATHEMATICS 2000 Les applications des math matiques an 2000 e par Mogens NISS
Aussi loin que remontent les id es denseignement et dapprentissage des e math matiques, certains aspects de leurs applications se sont retrouv s, explicitement e e ou implicitement, dans les programmes scolaires de la plupart des pays. Sil est vrai quavant le 20e si` cle les applications des math matiques concernaient principalement e e lenseignement primaire, elles ont par la suite fait leur chemin dans dautres niveaux, en parall` le avec le ph nom` ne de la massication de lenseignement secondaire et poste e e secondaire. Dans le cas de lenseignement secondaire, ce d veloppement sest produit e au cours de la p riode 19001960. Il sest poursuivi au niveau post-secondaire depuis e les ann es 1960, de sorte que les applications des math matiques sont maintenant e e pr sentes au premier cycle universitaire un peu partout dans le monde. e Ce texte porte sur le d veloppement des applications des math matiques durant le e e dernier tiers du 20e si` cle. Au cours de cette p riode, les tendances g n rales que nous e e e e venons de mentionner etaient pr sentes, bien que sous des formes nouvelles et dans un e contexte nouveau. Un aspect caract ristique en est que laccent sur certaines branches e des math matiques, habituellement regroup es sous le vocable de math matiques e e e appliqu es, fut remplac par la notion g n rale dapplication des math matiques (de e e e e e ` e quelque sorte que ce f t). De plus, la d marche ne fut pas restreinte a l tude de u e quelques exemples tri s sur le volet dapplications toutes cuites de facon typique, e sous la forme de mod` les standard mais fut elargie de mani` re a englober tout le e e ` processus de mise en uvre des math matiques dans des contextes extra-math matiques. e e Le processus fut bient t connu sous le nom de mod lisation ou de construction de o e mod` les (ou encore, dans le jargon scolaire, de r solution de probl` mes r els). e e e e On peut sommairement identier cinq raisons, somme toute assez diff rentes, pour e ` accorder un r le explicite aux applications et a la mod lisation dans le curriculum o e math matique. Ces raisons se retrouvent dans les arguments invoqu s en faveur des e e applications et de la mod lisation. A savoir : largument utilit (les applications e e et la mod lisation doivent faire partie des programmes an de rendre les etudiants e
272
M. NISS
` e capables dutiliser les math matiques dans dautres domaines, a l cole ou en dehors) ; e largument comp tence critique (il sagit de fournir aux etudiants les comp tences e e leur permettant de comprendre et dappr cier les emplois des math matiques signicatifs e e sur le plan social) ; largument formation (cest un outil efcace pour favoriser le d veloppement g n ral, chez lindividu, dattitudes, dhabilet s et de comportements en e e e e r solution de probl` mes) ; largument perception des math matiques (lapplication des e e e math matiques dans des domaines extra-math matiques repr sente une caract ristique e e e e importante des math matiques, et la mod lisation constitue justement le v hicule e e e par lequel passe lapplication) ; largument apprentissage des math matiques (les e ` applications et la mod lisation renforcent a la fois lapprentissage des math matiques e e et la motivation pour leur etude). e ` Ces raisons ont et avanc es par diff rents d fenseurs, a diff rentes epoques et e e e e avec des accents diff rents. Le mouvement des math matiques modernes et ses e e retomb es furent lobjet dattaques par des groupes inuents dans la soci t , car les e ee ` programmes r vis s ne fournissaient pas des dipl m s aptes a utiliser ce quils avaient e e o e appris. Cette critique, en grande partie fond e sur largument utilit , n tait pas le e e e propre des employeurs et des praticiens hors du monde de l ducation, puisquelle etait e aussi partag e par les math maticiens uvrant dans le domaine de lindustrie ainsi e e que par certains enseignants et didacticiens des math matiques. La philosophie souse jacente etait que pour permettre aux gens de d velopper des comp tences en rapport e e avec les applications et, surtout, avec la mod lisation, il fallait les leur enseigner. Au e cours des ann es 70 et au d but de la d cennie suivante, la port e de largument e e e e utilitaire fut elargie aux qualications concernant la vie de tous les jours et m me e la citoyennet . Pour certains educateurs, cela d boucha sur largument comp tence e e e critique. Vers 1970, un autre groupe denseignants et de didacticiens des math matiques e conjugua ses efforts avec certains utilitaristes. Non seulement les maths modernes ne satisfaisaient-elles pas les attentes des d fenseurs de lutilitarisme, mais une majorit e e d tudiants eprouvaient m me des difcult s a affronter les math matiques quon leur e e e ` e enseignait. En mettant laccent sur le r le-cl de la math matisation de la r alit dans o e e e e la compr hension de ce que sont les math matiques, ces enseignants et didacticiens e e privil giaient largument apprentissage des math matiques. Largument perception e e des math matiques, qui visait surtout les niveaux secondaire et post-secondaire, e commenca a emerger au cours des ann es 70 et 80, tandis que largument formation ` e connut son apog e au cours de la d cennie 70. e e Il est egalement possible de distinguer diverses phases en ce qui concerne le d veloppement des programmes et leur r alisation dans la pratique scolaire. Au cours e e de la premi` re phase, durant les ann es 70, laccent etait mis sur la d fense m me e e e e des applications et de la mod lisation, soutenues par la conception et la pr sentation e e de banques de mod` les et dapplications ainsi que par le compte rendu dexp riences e e ` ` denseignement, a petite et a grande echelle. Au cours de la d cennie suivante, lattention e porta sur les besoins des etudiants engag s dans de v ritables activit s de mod lisation, e e e e toutes les etapes du processus de mod lisation devant etre prises en compte de facon e ` s rieuse. La d cennie 90, quant a elle, na pas vu de d veloppements fondamentalement e e e nouveaux, les applications et la mod lisation ne repr sentant plus un sujet chaud. e e Ce nest cependant quau cours de cette d cennie que la recherche didactique sur les e applications et la mod lisation a vraiment pris son envol. Mais pour ce qui est de la e r alit de la salle de classe normale, cest une tout autre histoire : il semble en effet e e que ne sy soit jamais vraiment etablie une solide tradition dactivit s ayant trait aux e ` applications et a la mod lisation. e
APPLICATIONS OF MATHEMATICS 2000 by Mogens NISS
1. INTRODUCTION As long as there has been mathematics education, some aspects of the applications of mathematics have been present in the curriculum, explicitly or implicitly, in almost all countries. In former times, i.e. before the 20 th century, this pertained particularly to primary education. When higher educational levels became widely accessible to the general populace, the application of mathematics made its way into the agenda of mathematics education for these higher levels. This happened rst to secondary education, during the period 19001960. Then this development gradually gained momentum at the tertiary level since 1960 and is now rather manifest at the undergraduate level in most places of the world. This should come as no surprise. Basically, when mathematics education is being supplied to recipients who are not studying mathematics in order to become involved in its production (as researchers) or reproduction (as teachers), mathematics education will necessarily have to be justied by reference to matters that are extraneous to mathematics itself. Traditionally, such justication has consisted in referring either to the value mathematics has for the formation of the character and personality of the individual, or to the usefulness of mathematics to extra-mathematical life in everyday practice, on the job, in the community, or in society. The expansion of the educational system since the 60s, so as to make it cater for new and larger groups of students. has led to exactly this. Mathematics has been given, at still higher levels, to more and more students who will not primarily be preoccupied with the production or reproduction of mathematics as a discipline.
274
M. NISS
While this development has put the application of mathematics on the agenda of new (higher) educational levels, this does not imply that the application of mathematics has been given constantly increasing weight in the actual teaching and learning of mathematics. Throughout the 20 th century one can actually discern oscillations between utilitarian periods, during which the relevance and usefulness of mathematics education to extra-mathematical life have been emphasised, and puristic periods during which the intrinsic values of mathematics as part of human culture, or its signicance to the formation of the personality, are in the centre [Niss 1981 ; 1996]. However, these oscillations have been superimposed on the general trend already mentioned. So, although the role and importance of applications and modelling are not increasing in a literal sense they are, in fact, so on average. This paper concentrates on the development with respect to the application of mathematics in the last third of the 20th century. Within that period, too, the general trends outlined above are present, even though they have new backgrounds and take new forms during the period. It is a characteristic feature of the development during these three decades that the focus on certain branches of mathematics which were granted the status of applied mathematics was soon superseded by the general notion of applying mathematics (of whatever sort). This, in turn, was not conned to studying (samples of) particular, ready-made applications of mathematics typically in the form of standard models but became extended so as to encompass the entire process of putting mathematics to use towards matters extra-mathematical. This process became known under the label of modelling or model building ; in school contexts also sometimes as real-world problem solving. In order to generate a working terminology, let us use the shorthand applications and modelling for the whole package, including applied mathematics, applications of mathematics, mathematical models, and mathematical modelling, although there are important differences between these elements to keep in mind. The shift of emphasis from the former to the latter during the period we are considering will be further discussed below.
2. JUSTIFICATION
OF APPLICATIONS AND MODELLING IN THE CURRICULUM
In essence one can identify ve quite different reasons for assigning a manifest role to applications and modelling in a mathematics curriculum. These reasons are reected in the arguments put forward in the mathematics education literature in favour of applications and modelling [Blum & Niss
APPLICATIONS OF MATHEMATICS 2000
275
1991]. To avoid misunderstandings, these arguments are not (necessarily) my arguments, but arguments detected in the literature (cf. also [UNESCO 1973, 8182] and [Kaiser-Messmer 1986]). The order of presentation is neither chronological nor a manifestation of the popularity or priority assigned to each argument. (a) The utility argument. Ultimately, the majority of students of mathematics will be users rather than producers or reproducers of mathematics. Since experience shows that the ability to use mathematics is not an automatic consequence of knowing or even mastering pure mathematics, this ability has to be taught. So, applications and modelling should be included in the mathematics curriculum (at least for the majority of students) in order to enable them to use mathematics in other elds inside or outside of school. (b) The critical competency argument. It is a fact that mathematics is being widely used in society, actually to such an extent that it contributes to the shaping of society, for better and for worse. Multitudes of decisions and actions of great societal and social signicance are being taken under the involvement of mathematics. Whether such use is justied or not, it requires both mathematical and application and modelling competencies to understand and judge it. So, applications and modelling should be included in the mathematics curriculum in order to equip its recipients with a critical competency that allows them to understand and judge socially signicant uses of mathematics. (c) The formative argument. It is a major goal of mathematics education to help the formation of active, autonomous, creative, and exible individuals. One aspect of this is the ability to identify, pose, formulate, handle, and ultimately solve problems that one encounters on ones way. Sometimes such problems may involve the use of mathematics, not infrequently in a nonroutine setting. Sometimes they dont. However, in either case applications and modelling are an efcient tool for fostering general problem solving attitudes, abilitites and behaviours in individuals. So, this is a reason for including applications and modelling in the mathematics curriculum. These three arguments have one thing in common : they all focus on what applications and modelling can do for preparing students to master life outside of mathematics. Since, of course, mathematics is a necessary (but not sufcient !) prerequisite to applications and modelling, if one or
276
M. NISS
more of these arguments prevail they also lead to considering the issue what mathematics (education) can do for applications and modelling, and by transitivity, what mathematics (education) can do for life outside of mathematics. The following two arguments, in contrast, have the opposite orientation, in that they bring answers to the question : What can applications and modelling do for mathematics, and its teaching and learning ? (d) The picture of mathematics argument. Let us assume that mathematics education is meant to provide its recipients with a comprehensive, balanced, and representative picture of mathematics as a discipline and of its position and role in the world. As the application of mathematics to extramathematical areas is indeed a characteristic feature of mathematics, and as modelling is the vehicle through which the appplication of mathematics is brought about, applications and modelling should occupy a manifest position in the mathematics curriculum. (e) The learning of mathematics argument. Numerous mathematical entities (including objects, phenomena, relations, and concepts) are to a large extent linked to corresponding entities in the real world. Oftentimes mathematical entities are abstractions derived from attempts to model aspects of this world. So, in order really to understand these entities it is necessary or at least very helpful to consider their origin in modelling. Moreover, since most mathematical concepts, methods and results are actually applicable to dealing with matters outside of mathematics, working with applications and modelling serves to underpin and scaffold the understanding and appreciation of mathematics. So, both the learning of mathematics and the motivation for studying it are strengthened by giving applications and modelling a manifest position in the mathematics curriculum.
3. TRENDS 19672000 The reasons outlined above are given different emphases by different quarters at different times. The era of the new (or modern) maths has been dealt with, in this symposium, by H l` ne Gispert. Sufce it here to emphasise that societys ee overarching reason for supporting curriculum reform during the 60s was clearly utilitarian. Society was convinced that mathematical knowledge was crucial to a rapidly growing portion of occupations which required independent,
277
exible, creative, and efcient employees able to activate their mathematical knowledge in non-routine situations originating in their work [Cooper 1985, chap. 58]. As a matter of fact, the reformers behind the new maths movement agreed completely [OECD 1961, 107]. They just found that nothing mathematical was so useful and applicable as good, pure, general mathematical structures (e.g. logic, sets, relations, groups). By virtue of their generality these structures were not tied to specic contexts but were, on the contrary, of potential universal relevance to all sorts of elds from the world outside of mathematics. It also belongs to this part of the story that it was the new maths movement that was instrumental in including probability and statistics in the secondary school curriculum because of the signicance of these topics to the application of mathematics. Employers in general, and industrialists in particular, have always insisted on the usefulness and applicability of mathematics as the ultimate justication for its prominent position in the educational system. How the educational system will achieve that the people it produces possess the knowledge and skills needed to use mathematics in the work place is, in principle, of secondary importance as long as it happens. However, it soon became clear that the new maths reform did not keep its promises by producing the sorts of people that were in demand by employers. Graduates from the reformed curricula turned out not to be any better at using mathematics for practical purposes than was the case before. One could even say that they were worse in this respect because now many of them could not even do the computations, say in arithmetic, they used to be able to do. It is true that the reformed curricula did produce graduates from secondary and tertiary education who had beneted greatly from the new approach, in terms of pure mathematical knowledge and insight. But this did not automatically enable them to make use of what they had learnt. So, the results of the new maths came under harsh attack by large and inuential quarters in society (for an example, see [Hammersley 1968]). This provided one essential stimulus for upgrading the role of applications and modelling in mathematics curricula at virtually all levels. Needless to say, this line of reasoning was largely based on the utility argument. It was shared not only by employers and practitioners outside of the mathematics education community but also by increasingly strong and vocal groups of mathematicians [Pollak 1968 ; 1976 ; Klamkin 1968], some of whom worked in industry, and mathematics educators [Ford & Hall 1970], primarily in the Anglo-Saxon world, but also in Germany, the Netherlands, and Scandinavia.
278
M. NISS
The underlying philosophy, based on experience, was (and is) that if you want people to obtain competence in applications and modelling you have to teach it to them. This goes for all the aspects of the applications and modelling package, but above all for modelling. To become able to carry through the entire modelling process from problem identication and specication, through within-model problem solving, to model validation, and including all the steps in between, requires both substantial training and multiple experiences, which in turn have to be obtained by means of special activities. It is according to those who put forward the utility argument the responsibility of mathematics education to take care of such training and to provide such multiple experiences, in order ultimately to ensure that the applications and modelling competency thus formed has a sufciently wide and general scope. If circumstances allow for it, collaboration with other subjects, such as physics, biology, economy, technology, is certainly most valuable too, but only in mathematics is it possible to generate and display a broad spectrum of applications and modelling experiences taken from different elds. Along the road, during the 1970s and early 80s, the scope of the utility argument was expanded from focusing mainly on job qualications to emphasising also peoples everyday life and citizenship. This happened concurrently with the political and societal development that took place in that decade, which led to a strong emphasis in several quarters on the sociopolitical relevance of education. The well-known Cockcroft report [Cockcroft 1982] in the UK gave particular emphasis to the need of preparing people in general for everyday life and citizenship. For some mathematics educators, especially but certainly not exclusively in Germany and Scandinavia (e.g. Damerow et al. [1974], Mellin-Olsen [1987], Skovsmose [1994], and Volk [1979], amongst many others), this gave rise to the critical competency argument which at rst sight may be perceived as a variant or at least an offspring of the utility argument, although its focus is in fact rather different. Around 1970 a different group of mathematics educators joined forces with some of the utilitarians. Not only did the new maths fail to full the expectations or demands of utilitarian quarters ; it also turned out that a major proportion of students had severe difculties in coming to grips with the mathematics taught to them as such, whether or not an applications and modelling perspective was present. To many students, mathematics appeared to be a completely abstract game played according to some random set of rules coming out of the blue. Perhaps the game could sometimes be good fun, to some students, but it was played in splendid isolation from the rest of the world.
279
This state of affairs was observed with great concern by a number of mathematics educators, amongst whom Hans Freudenthal was a protagonist. Already in 1967 an international colloquium, sponsored by ICMI, the IMU, and the Dutch government, was held in Utrecht under the title How to teach mathematics so as to be useful. It gathered a varied group of mathematicians and mathematics educators who insisted on the need to take the application of mathematics seriously. The proceedings of that colloquium were published in the rst volume of Educational Studies in Mathematics in 1968. Freudenthals opening address deserves to be quoted in part :
Since mathematics has proved indispensable for the understanding and the technological control not only of the physical world but also of the social structure, we can no longer keep silent about teaching mathematics so as to be useful. [] The large majority of students are not able to apply their mathematical classroom experiences, neither in the physics or chemistry school laboratory nor in the most trivial situations of daily life. [] In its rst principles mathematics means mathematizing reality, and for most of its users this is the nal aspect of mathematics, too. [] If we do not succeed in teaching mathematics so as to be useful, users of mathematics will decide that mathematics is too important a teaching matter to be taught by the mathematics teacher. Of course this would be the end of all mathematics education. [Freudenthal 1968, 48]
While this quotation primarily emphasises the utility of mathematics, Freudenthal in his address also touched upon the key role of mathematisation of reality in understanding what mathematics is all about, and in coming to grips with its concepts and methods. Thus he and other participants in the colloquium (e.g. Krygowska [1968]) also invoked and emphasised the learning of mathematics argument. This is an argument which has always been strong with mathematics educators and (some) university mathematicians, whereas it has never attracted much attention in quarters outside of the educational system. The same is true with the picture of mathematics argument which began to emerge amongst some (but rather few) mathematics educators and mathematicians in the 70s and 80s, primarily addressing secondary and tertiary levels. As nally regards the formative argument, mainly put forward by mathematics educators, this had its heydays in the 70s, but then lost some of its momentum along with the growth of a general distrust of transfer in the 80s and 90s.
280 4. APPLICATIONS
M. NISS
AND MODELLING IN THE CURRICULUM
So much about the arguments for applications and modelling in mathematical education. How about the actual development of curricula and implementation in classroom practice ? Here it is possible to discern different phases. In the rst phase, in the 70s, the main emphasis was on the very advocacy of applications and modelling, which was supported by the generation and presentation of source collections of models and applications and by reporting teaching experiments in small or large scale, all of which were meant to serve two purposes. Firstly, to demonstrate that it is actually possible to nd examples or cases that are viable and accessible to the teaching of mathematics at various levels. Secondly, to serve as a bank of resources for teachers and textbook authors who wanted to include some element of applications and modelling in their teaching. In addition to establishing such resource banks, much work was put into recruiting particularly interested teachers to take up applications and modelling work in their own teaching practice. The establishment in 1983 in the USA of COMAP, the Consortium for Mathematics and its Applications, as an organisation to further the creation of curricula and teaching materials was a reection of this development [Garfunkel 1998]. However, in spite of the danger of oversimplifying the situation in the 70s, it seems fair to state that the majority of applications and modelling teaching round the world focused on ready-made models that were presented to students for further study. It was also in the 70s that applications and modelling began to gure in prominent places in the ICMEs. At ICME-3 in Karlsruhe in 1976, a full section was devoted to applications and modelling, and an often quoted report by Henry Pollak (with the slightly misleading title The interaction between mathematics and other school subjects) was published by UNESCO [1979, 232248]. Also at ICME-4 in Berkeley, 1980, applications formed a major theme, reected in a chapter of 70 pages in the Proceedings [Zweng et al. 1983, no. 8]. In the 80s attention was drawn to the need for students to engage in real modelling activities, where all the stages of modelling were to be taken seriously, also those that cannot be labelled as predominantly mathematical. Since this is time consuming, such work was found primarily in experimental or innovative settings characterised by a great deal of enthusiasm on the part of those involved. Quite a few large scale projects in applications and modelling were established in the 80s.
281
At the level of countries, there is little doubt that the UK had a leading role in emphasising modelling [Ford & Hall 1970 ; Burghes 1980]. Various groups such as the Spode Group were formed to promote and develop modelling activities in school and university. In many of the new polytechnics in Britain a lot of work was done in this direction. It is also on the initiative of British mathematics educators that the series of International Conferences on the Teaching of Mathematical Modelling and Applications (the ICTMAs) was instigated, the rst one being held in Exeter in 1983 [Berry et al. 1985]. Since then an ICTMA has been held every odd year [Berry et al. 1986 ; Blum et al. 1989 ; Niss et al. 1991 ; de Lange et al. 1993 ; Sloyer et al. 1995 ; Houston et al. 1997 ; Galbraith et al. 1998] with ICTMA-11 going to be held in Milwaukee in 2003. As the scope of these conferences became widened from tertiary applications and modelling, lower educational levels were taken on board as of the late 80s. A few years ago, the informal community behind the ICTMAs decided to formalise itself into the International Community for the Teaching of Mathematical Modelling and Applications [www.infj.ulst.ac.uk/ictma/]. At ICTMA-10 in Beijing 2001, this community further decided to apply for the status of Afliated Study Group of ICMI. The way in which applications and modelling were included in the ICMEs of the 1980s (1980 not included) was changed to give more emphasis to the modelling aspect. This happened along with the new format adopted for the ICMEs initiated at ICME-5 in Adelaide (1984). At that congress, Theme Group 6 was devoted to applications and modelling [Carss 1986, 197 211]. At ICME-6 in Budapest (1988), two theme groups dealt with matters pertaining to applications and modelling, no. 3 problem solving, modelling and applications, and no. 6 mathematics and other subjects [Hirst & Hirst 1988]. Based on the work of these two theme groups a joint book was subsequently published [Blum et al. 1989]. In the 90s one can hardly say that fundamentally new developments occurred. The trends of the 70s and 80s were consolidated in some places (for an overview, see [de Lange 1996]), whereas they lost steam in others. In general it seems fair to claim that applications and modelling was no longer such a hot topic in the 90s. Indeed, it continued to be cultivated in ICTMA circles, which were also expanded in the 90s. Applications and modelling were given a status quo position in ICME-7 in Qu bec, except that modelling was emphasised e even more than in the past. Working Group 14 dealt with modelling in the classroom [Breiteig et al. 1993]. In ICME-8 in Seville, one could discern a weakening of the focus, in that applications and modelling tended to be spread over a larger set of groups. Thus, one of the Working Groups had
282
M. NISS
the title linking mathematics with other school subjects, while one of the Topic Groups was called problem solving throughout the curriculum. No group specically addressed applications and modelling as such. At ICME-9 in Tokyo/Makuhari 2000, applications and modelling re-entered the congress programme in the Topic Study Group (no. 9) on mathematical modelling and links between mathematics and other subjects. It should be added that it was only in the 90s that didactical research into applications and modelling gained some momentum [Niss 2001]. As a reection of this momentum, ICMI has decided to mount an ICMI Study on the teaching and learning of mathematical applications and modelling. One thing is what happens at the level of the didactics of mathematics in terms of discussion, research and development. Quite a different thing is what really happens in the mainstream classroom. Although I have no solid data to offer, it does seem from reports of protagonists in the eld that substantial activity in applications and modelling was never secured an unchallenged foothold in such classrooms. Some attention appears to be paid in most places, but serious emphasis is conned to special places and institutions.
REFERENCES BERRY, J. S., I. D. HUNTLEY, D. J. G. JAMES and A. O. MOSCARDINI (eds). Teaching and Applying Mathematical Modelling. Ellis Horwood, Chichester, 1985. BERRY, J. S., D. N. BURGHES, I. D. HUNTLEY, D. J. G. JAMES and A. O. MOSCARDINI (eds). Mathematical Modelling Methodology, Models and Micros. Ellis Horwood, Chichester, 1986. BLUM, W., J. S. BERRY, R. BIEHLER, I. D. HUNTLEY, G. KAISER-MESSMER and L. PROFKE (eds). Applications and Modelling in Learning and Teaching Mathematics. Ellis Horwood, Chichester, 1989. BLUM, W. and M. NISS. Applied mathematical problem solving, modelling, applications, and links to other subjects state, trends, and issues in mathematics instruction. Educational Studies in Mathematics 22 (1991), 3768. BLUM, W., M. NISS and I. HUNTLEY (eds). Modelling, Applications and Applied Problem Solving Teaching Mathematics in Real Context. Ellis Horwood, Chichester, 1989. BREITEIG, T., I. HUNTLEY and G. KAISER-MESSMER (eds). Teaching and Learning Mathematics in Context. Ellis Horwood, Chichester, 1993. BURGHES, D. N. Mathematical modelling : a positive direction for the teaching of applications of mathematics at school. Educational Studies in Mathematics 11 (1980), 113131. CARSS, M. (ed.) Proceedings of the Fifth International Congress on Mathematical Education. Birkh user, Boston, 1986. a
283
COCKCROFT, W. H. Mathematics Counts. Report of the Commission of Inquiry into the Teaching of Mathematics in Schools. Her Majestys Stationery Ofce, London, 1982. COOPER, B. Renegotiating Secondary School Mathematics. A Study of Curriculum Change and Stability. The Falmer Press, London and Philadelphia, 1985. DAMEROW, P., U. ELWITZ, C. KEITEL and J. ZIMMER. Elementarmathematik : Lernen f r die Praxis ? Ernst Klett Verlag, Stuttgart, 1974. u FORD, B. and G. G. HALL. Model building an educational philosophy for applied mathematics. International Journal of Mathematical Education in Science and Technology 1 (1970), 7783. FREUDENTHAL, H. Why to teach mathematics so as to be useful ? Educational Studies in Mathematics 1 (1968), 38. GALBRAITH, P., W. BLUM, G. BOOKER and I. HUNTLEY (eds). Mathematical Modelling. Teaching and Assessment in a Technology-Rich World. Ellis Horwood, Chichester, 1998. GARFUNKEL, S. A sojourn in mathematics education. In : C. Alsina et al. (eds.), 8th International Congress on Mathematical Education Selected Lectures, 215 219. S.A.E.M. Thales, Sevilla, 1998. HAMMERSLEY, J. M. On the enfeeblement of mathematical skills by the teaching of modern mathematics and similar intellectual trash in schools and universities. Bulletin of the Institute of Mathematics and Its Applications 4 (1968), 6885. HIRST, A. & K. (eds). Proceedings of the Sixth International Congress on Mathematical Education. J nos Bolyai Society, Budapest, 1988. a HOUSTON, S. K., W. BLUM, I. HUNTLEY and N. T. NEILL (eds). Teaching and Learning Mathematical Modelling. Albion Publishing, Chichester, 1997. KAISER-MESSMER, G. Anwendungen im Mathematikunterricht. Band 12. Verlag Franzbecker, Bad Salzdetfurth, 1986. KLAMKIN, M. On the teaching of mathematics so as to be useful. Educational Studies in Mathematics 1 (1968), 126160. KRYGOWSKA, A. Z. Processus de la math matisation dans lenseignement. Educational e Studies in Mathematics 1 (1968), 916. DE LANGE, J. Using and applying mathematics in education. In : International Handbook of Mathematics Education, Vol. 1 (eds A. Bishop et al.), 4997. Kluwer, Dordrecht, 1996. DE LANGE, J., I. HUNTLEY, C. KEITEL and M. NISS (eds). Innovation in Mathematics Education by Mathematical Modelling and Applications. Ellis Horwood, Chichester, 1993. MELLIN-OLSEN, S. The Politics of Mathematics Education. Reidel, Dordrecht, 1987. NISS, M. Goals as a reection of the needs of society. In : Studies in Mathematics Education, Vol. 2 (ed. Morris, R.), 121. UNESCO, Paris, 1981. Goals of mathematics teaching. In : International Handbook of Mathematics Education, Vol. 1 (eds A. Bishop et al.), 1147. Kluwer, Dordrecht, 1996. Issues and problems of research on the teaching and learning of applications and modelling. In : Modelling and Mathematics Education ICTMA 9 : Applications in Science and Technology (ed. J. F. Matos, W. Blum, S. K. Houston and S. P. Carreira), 7288. Horwood Publishing, Chichester, 2001.
284
M. NISS
NISS, M., W. BLUM and I. HUNTLEY (eds). Teaching of Mathematical Modelling and Applications. Ellis Horwood, Chichester, 1991. OECD. New Thinking in School Mathematics. OECD, Paris, 1961. POLLAK, H. On some of the problems of the teaching applications of mathematics. Educational Studies in Mathematics 1 (1968), 2430. What industry wants a mathematician to know and how we want them to know it. Educational Studies in Mathematics 7 (1976), 109112. SKOVSMOSE, O. Towards a Philosophy of Critical Mathematics Education. Kluwer, Dordrecht, 1994. SLOYER, C., W. BLUM and I. D. HUNTLEY (eds). Advances and Perspectives in the Teaching of Mathematical Modelling and Applications. Water Street Mathematics, Yorklyn (Delaware), 1995. UNESCO. New Trends in Mathematics Teaching, Vol. III. UNESCO, Paris, 1973. New Trends in Mathematics Teaching, Vol. IV. UNESCO, Paris, 1979. VOLK, D. (ed.) Kritische Stichw rter zum Mathematikunterricht. W. Fink Verlag, o M nchen, 1979. u ZWENG, M., T. GREEN, J. KILPATRICK, H. POLLAK and M. SUYDAM. Proceedings of the Fourth International Congress on Mathematical Education. Birkh user, Boston, a 1983.
REACTION by Gerhard WANNER
Even though I have coauthored a textbook on analysis [Hairer & Wanner 1995], I see myself as a specialist in numerical analysis, not as an expert in mathematics education. Like others, I was a pupil and a student myself some 40 years ago. Since then I have taught mathematics at various levels to students of computer science, physics, and mathematics. But, after the three brilliant lectures we have heard, I have the awkward feeling of re-experiencing Galileos dialogue between two intelligent participants (Salviati, Sagredo) and a stupid one (Simplicio) with the only improvement that here we have three intelligent ones against one stupid. All the texts mentioned, throughout the century, stress the importance of applied mathematics in teaching. My overall impression is that they differ mainly in what is meant, in each period, by applied mathematics : notion of function from algebraic and geometric viewpoints for computations of modern industrial applications, use of 4-decimal logarithms in the early century ; probability and stochastics, operational research, numerical analysis in the years 19501970 ; much of this increased interest in applications came from industrial research labs in the sequel of World War II and after the sputnik shock ; applications and modelling of large real world problems in the third period. Again and again the importance of teaching applied mathematics has been emphasized, not only for preparing students to master their future life outside of mathematics, but also for improving the learning of mathematics itself.
286
G. WANNER
TABLE 1 Level of acquaintance with high-school mathematics (beginning rst-year students in Geneva, October 1985)
Faculty :
Bio PhM Info Geo Chm Eco GE For
1. A cylinder of revolution with radius r and altitude h 70% 86% 63% 81% 82% 37% 56% 43% has surface area A 2. Compute, without using a calculating machine : Log 5 Log 20 Log (Log denotes logarithm to base 10)

32% 60% 57% 6% 60% 17% 29% 29%
3. One throws two dice simultaneously. Give the proba- 49% 50% 33% 50% 55% 35% 41% 35% bility that the sum of numbers on their sides is equal to 3 4.
y 0 2 x
Give the equation of the drawn circle. 27% 47% 31% 18% 49% 11% 22% 21%
as x
6. The set a b c contains a pair of orthogonal vectors. Which one ?

2 1 3
, b
1 5 1
, c
3 2 5
7. Compute the derivative of the function f (x) x f (x)
8. Compute
1 2 0 x
dx
Students with 7 or 8 correct answers : Students with 0 or 1 correct answer :
Bio = Medicine, Biology, Pharmacy; PhM = Physics, Mathematics; Info = Computer Science; Geo = Geology; Chm = Chemistry; Eco = Economics and Social Sciences; GE = Global percentage for Geneva; For = Foreign students and students with incomplete secondary schooling
5. To what value does the function f (x)
1 1 x2
tend
90% 93% 92% 75% 95% 84% 87% 85%

38% 64% 49% 37% 75% 26% 36% 28%
and 81% 88% 84% 93% 86% 57% 71% 67%
74% 88% 72% 62% 80% 44% 60% 53% TOTAL 58% 72% 60% 53% 73% 39% 50% 45% 14% 43% 29% 12% 35% 5% 14% 10% 4% 3% 2% 0% 2% 18% 10% 16%
APPLICATIONS OF MATHEMATICS
287
ACTUAL FACTS IN SCHOOLS. Despite these admirable resolutions and discussions about the didactics of mathematics, research and development, it is a striking reality that in mainstream classrooms one can hardly nd any serious activity in applied mathematics. This concluding remark of M. Niss contribution (Although I have no solid data to offer [Niss 2003, 282]) is also supported by my own personal experience. Indeed, I took part in a commission set up by the Rectors of all Swiss universities in 1985. The object was to study the acquaintance with high-school mathematics of students entering the universities. All rst-year students in science, medicine, and economics were given an anonymous questionnaire with eight questions of an extremely simple nature (simple as compared to the usual curriculum). The results for Geneva are displayed in Table 1. Among the various questions, the more elementary and the more applied they were (like the probability quiz), the worse were the results. At the other extreme, abstract limits, derivatives, and integrals were known best. Another episode in the same direction was David Mumfords astonishment when he
tried (unsuccessfully) to get each high school in which my children were enrolled to go outside during geometry and nd out how tall the oak in the yard really is. Instead they buckled under to the educational establishment [Mumford 1999, 5]
WHAT
COULD BE THE REASONS FOR THIS DISCREPANCY ?
In the next paragraphs we suggest three principal reasons which can be invoked to analyse this phenomenon. PURE MATHEMATICS HAS HIGHER REPUTE. There are numerous statements by pure mathematicians (the true mathematicians of Godement) against applications :
Les sectateurs des math matiques appliqu es, de lanalyse num rique et de e e e linformatique font preuve dans toutes les universit s du monde de tendances e imp rialistes beaucoup trop manifestes pour que les vrais math maticiens 1 ) e e [Godement 1998, IX]
This tradition, which is absent from the work of Newton, Euler or Gauss, goes back, as Felix Klein explains, to the times of Abel and Jacobi :
1 ) The partisans of applied mathematics, numerical analysis, and computer science display, in all the worlds universities, imperialist tendencies which are much too manifest for true mathematicians to
288
G. WANNER
durch die Gr ndung des Journals f r die reine und angewandte Mathematik u u (1826). Nimmt man heute einen Band dieser Zeitschrift zur Hand, so mag der Titel vielleicht Verwunderung erregen. [] Das neuhumanistische Ideal der reinen Wissenschaft als Selbstzweck, das die Verachtung aller Nutzlichkeit im gemeinen Sinne in sich barg, f hrte bald zu einer geissentlichen Abkehr von u allen der Praxis zugewandten Bestrebungen [] und stempelte es zu einem Organ abstrakter Spezialmathematik von strengster Auspr gung, die ihm den a Scherznamen Journal f r reine, unangewandte Mathematik eingetragen hat. 2 ) u [Klein 1926, 9495]
Future high-school teachers learn this frame of mind during their studies at university.
Nous navons pas besoin dordinateur. 3 ) [G. de Rham, to the Dean of the Faculty, Geneva, 1970] Les math matiques pures, cest comme un tableau de Rembrandt ; les e math matiques appliqu es, cest comme un peintre en b timent. 4 ) [A member e e a of the Maths Department in Geneva, around 1975]
Not surprisingly, they are badly prepared for applications.

Ich habe immer wieder beobachtet, da Mathematiker und Physiker mit ab geschlossenem Examen uber theoretische Ergebnisse sehr gut, aber uber die einfachsten N herungsverfahren nicht Bescheid wuten. 5 ) [Collatz 1951, V] a
APPLIED MATHEMATICS IS SEEN ONLY AS UTILITARIAN. Even the title of this chapter Applications of mathematics : mathematics as a service subject makes us believe that the only role of applied mathematics is to be useful : something served by noble servants in dress coat to physicists, chemists, biologists sitting there as the clientele, after being prepared in la haute cuisine by pure mathematicians. It appears widely forgotten that the other way round applications often had an enormous inuence on the development of pure mathematics itself.
2 ) by founding the Journal for pure and applied mathematics (1826). If today we browse through a volume of this publication, we may well nd the title astonishing. [] The neohumanistic ideal of pure science for its own sake, which implied the contempt of every form of utility in the ordinary sense, soon led to deliberately turning away from all aspirations oriented towards practice. [] This hallmarked it as an organ of abstract special mathematics of the strongest mintage and earned it the humorous name of Journal for pure, unapplied mathematics. 3 ) We have no need for computers. 4 ) Pure mathematics is like a painting by Rembrandt; applied mathematics is more like a house-painters job. 5 ) I have repeatedly observed that mathematicians and physicists, upon completion of their exams, knew a lot about theoretical results but nothing about the simplest approximation procedures.
289
History is rich in such examples, beginning with Keplers astronomy used for the discovery of Newtons gravitation law and of differential calculus ; the brachystochrone problem leading to variational calculus (this will be further developed in the next paragraph) ; astronomical calculations by Euler and Gauss :
In 1735 the solving of an astronomical problem, proposed by the Academy, for which several eminent mathematicians had demanded some months time, was achieved in three days by Euler with aid of improved methods of his own. [] With still superior methods this same problem was solved later by K. F. Gauss in one hour ! [Cajori 1893 (1980), 232]
Another example is Fouriers theory of heat, which led to Dirichlets concept of function, Riemanns integral, Cantors set theory and Sturm-Liouvilles spectral theory in Hilbert spaces :
Jajouterai que le livre de Fourier a une importance capitale dans lhistoire des math matiques et que lanalyse pure lui doit peut- tre plus encore que lanalyse e e appliqu e. 6 ) [Poincar 1895, 1] e e
Further, Poincar s M canique c leste led to symplectic geometry and KAMe e e theory ; Curtiss and Hirschfelders chemical calculations led to the theory of stiff differential equations, and so forth. We may also mention logarithms, which were invented in the early 17 th century for the sake of practical calculations and have been used for several centuries. After pocket calculators came into use, logarithms were thought to be superuous for school teaching, and a whole generation of students never saw them in high school. The enormous importance they have for all sciences was thereby totally overlooked. The same fate is happening now to descriptive geometry. APPLIED MATHEMATICS IS MORE DIFFICULT TO TEACH. The teaching (and understanding) of applied mathematics gives rise to additional difculties for the teacher and for the students, such as : managing computer tools, learning programming languages, tracking wrong answers due to rounding errors, knowing the application eld like physics or chemistry, working with physical constants and conversion factors, and also the difculty of translating the applied problem conveniently into a mathematical one. For example, it is easier to solve 5 8 3 x
6 ) I shall add that Fouriers book is of utmost importance in the history of mathematics and that it has been even more protable, perhaps, to pure than to applied analysis.

290 than to answer the question
G. WANNER
5 workers accomplish a certain task in 8 hours ; how many hours will 3 workers need ?
A famous example in the same vein (though more serious) concerns the brachystochrone problem. Its history is outlined in the following three quotations.
PROBLEMA NOVUM Ad cujus solutionem Mathematici invitantur. Datis in plano verticali duobus punctis A & B , assignare Mobili M viam AMB , per quam gravitate sua descendens, & moveri incipiens a puncto A , brevissimo tempore perveniat ad alterum punctum B . 7 ) [Bernoulli 1696]
Initially it was a problem of applied sciences, asking for the curve of fastest descent, which de lH pital could not solve : o
Ce probleme [me] paroist des plus curieux et des plus jolis [que] lon ait encore propos et je serois bien aise de my appliquer ; mais pour cela il seroit e ` necessaire que vous me lenvoyassiez reduit a la mathematique pure, car le phisique membarasse 8 ) [letter of de lH pital to Joh. Bernoulli, 15 June o 1696 : NFG 63, 319]
Only with the knowledge of a certain amount of physics did this become a problem of pure mathematics :
` Voycy donc maintenant comme le probleme se reduit a la pure mathematique : Dentre toutes les lignes AMB qui joignent deux points donn s A et B on e cherche la nature de celle dont
soit la plus petite. 9 ) [letter of Joh. Bernoulli to de lH pital, 30 June 1696 : o NFG 64, 321]
7 ) Let two points A and B be given in a vertical plane. Determine the path AMB that a point M must follow so that, starting from A, it reaches B in the shortest time under its own gravity. 8 ) This problem looks to me one of the most intriguing and one of the nicest that has ever been proposed, and I would much like to dedicate myself to it; but it would be necessary for this that you send it to me reduced to pure mathematics, since the physical [aspect] disturbs me 9 ) Here is now how the problem reduces to pure mathematics : From among all paths AMB which connect two given points A and B, one is looking for the nature of the one for which
is smallest.
dx2
dy2 y
dx2
dy2 y
291
Our last quotation, from Daniel Bernoulli, shows what difculties can arise the other way round for applied scientists.
Den Hrn. DAlembert halte ich f r einen grossen mathematicum in abstractis ; u aber wenn er einen incursum macht in mathesin applicatam, so horet alle estime bey mir auf : seine Hydrodynamica ist viel zu kindisch, dass ich einige estime f r ihn in dergleichen Sachen haben konnte. Seine pi` ce sur les vents will u e nichts sagen und wenn Einer alles gelesen, so weiss er so viel von den ventis, als vorhero. Ich vermeinte, man verlange physische Determinationen und nicht abstracte integrationes. Es f ngt sich ein verderblicher go t an einzuschleichen, a u durch welchen die wahren Wissenschaften viel mehr leiden, als sie avancirt werden, und w re es oft besser f r die realem physicam, wenn keine Mathematik a u auf der Welt w re. 10 ) [letter of Daniel Bernoulli to Euler, 26 January 1750 : a Fuss 56, 649650]
Another weak point with applied mathematics is that the subjects easily become boring, outdated, politically incorrect or ludicrous. For example, the above problem of the ve workers can have the avour of a big manager 3 x is perfectly exploiting underpaid workers, while the equation 5 8 neutral. The same impression is conveyed by the many applied examples in Eulers Algebra (Opera Omnia, vol. 1), where there is always someone who lends money or buys clothes in long forgotten currencies. It is also interesting to compare the applied articles in the rst issues of the Crelle Journal, which discuss some uninteresting motions of steam engine parts or water-works, with the pure contributions of Abel and Jacobi, which still today glitter in eternal youth and beauty.

REFERENCES BERNOULLI, Joh. Problema novum Mathematicis propositum. Acta Eruditorum Lips. 6 (June 1696), 269. (Opera omnia I, 161.) CAJORI, F. A History of Mathematics. Macmillan, New York, 1893 ; 3rd ed., Chelsea, New York, 1980. COLLATZ, L. Numerische Behandlung von Differentialgleichungen. Springer, Berlin, 1951. FUSS, P.-H. Correspondance math matique et physique de quelques c l` bres g om` tres e ee e e du XVIIIe si` cle. Tome II. Acad mie Imp riale des Sciences, St-P tersbourg, 1843. e e e e
10 ) I hold Mr. DAlembert for a great mathematicum in abstractis; but when he makes an incursum into mathesin applicatam, then I no longer have any estime : his Hydrodynamica is much too childish for me to have any estime for him in such matters. His pi` ce sur les vents e is meaningless and when one has read it all, one knows as much about the ventis as before. I imagined that one required physical determinations and not abstract integrationes. One begins to see a pernicious go t creeping in, from which the true sciences suffer much more than they advance. u Would it often not be better for the realem physicam if there were no mathematics on earth ?
292
G. WANNER
GODEMENT, R. Analyse math matique I. Springer, Berlin, 1998. e HAIRER, E. and G. WANNER. Analysis by its History. (Undergraduate Texts in Mathematics) Springer, New York, 1995. KLEIN, F. Vorlesungen uber die Entwicklung der Mathematik im 19. Jahrhundert. Teil I. Springer, Berlin, 1926. MUMFORD, D. Preface. In : H. M. Enzensberger, Drawbridge Up ; Mathematics A Cultural Anathema, 56, A. K. Peters, Natick, 1999. NFG NATURFORSCHENDE GESELLSCHAFT BASEL (ed.) Der Briefwechsel von Johann Bernoulli. Band I. Birkh user, Basel, 1955. a NISS, M. [2003] Applications of Mathematics 2000. This volume, 271284. POINCARE, H. Th orie analytique de la propagation de la chaleur. Lecons (189394). e Carr , Paris, 1895. e
GENERAL DISCUSSION (reported by Chris WEEKS)
The discussion from the participants, following Gerhard Wanners reaction to the three talks, fell not surprisingly into the two broad topics of :
what meaning should be attached to the term applied mathematics ?
and
what does this imply for the mathematics curriculum ?
An aspect of the second proved to be of major importance, namely :

what is the role of assessment in the teaching of applied mathematics ?
The discussion did not fall neatly into just one of these categories as each had implications for the other. In fact, there was the view that the term applied mathematics was itself a construct of the teaching of mathematics. As one French participant put it :
The distinction between real and applied mathematics is not very clear. At the beginning of my career I was a pure mathematician ; now I am an applied mathematician, but personally I cannot see the difference. Certainly there is a difference, but it is made by those people who write pure or applied on their hats and maintain their independence ; we even see some who will not talk to one another. In my view it is not a matter of epistemology or didactics but une affaire humaine.
THE MEANING OF APPLIED MATHEMATICS. The meaning of applied mathematics, or what meaning should be given to the term, dominated the beginning of the discussion. Mathematics, where it is not undertaken as it were for its own sake, takes place within some exterior context and the mathematics then becomes applicable. Another term, now in constant use, is modelling. These ideas imply mathematics being used in some way to deal with some problem not present in the mathematics itself. But the question that then arises is
294
GENERAL DISCUSSION
whether or not the activity of applied mathematics is in reality physics or engineering or economics, for which mathematics is just one of the tools being used, or whether it is actually pure mathematics being carried out according to some given rules. Several speakers were of the view that the essential of applied mathematics lies in the translation of the real-world problem into a mathematical form. Once this mathematisation has been accomplished, the problem then becomes (simply) a problem of pure mathematics though the solution would need to be interpreted within the originating context. The example of LH pitals reaction to Bernoullis brachystochrone problem, given o by Gerhard Wanner, was found to be a useful reference example for the discussion. The real essence of applied mathematics is to simplify reality in order to carry out some mathematics on it. Even then, when the problem is solved, it is only rarely a reasonable approximation to the reality from which the problem derived. Of course there are occasionally some good mathematical models of the real world, but then the equations are often too difcult to solve (a good example is the differential equations describing uid dynamics). Maybe it would be better to think of modelling, or applied mathematics, as an enrichment (of context and ideas) and not as a separate subject. Mixed mathematics would be a better term. For example, for the study of abstract algebra earlier in the 20th century there was no mathematics book available, so it was necessary to turn to the physicists (who knew more about groups early in the 20th century than did professional mathematicians). Yet the theory of groups is now considered to be a topic belonging exclusively to pure mathematics. It is a constant fact in mathematics that mathematics comes from elsewhere, added one speaker. Another speaker pointed out that every person learning mathematics starts out from a situation where there is no distinction between pure and applied mathematics. Children just have questions puzzling them, what Freudenthal called phenomena. In this sense the distinction between pure and applied mathematics is quite articial. There was also the question of whether applied mathematics in some way showed that mathematics was useful (a theme that was picked up again in the discussion of the place of applied mathematics in the mathematics curriculum). But the question of the utility of mathematics begs a question. Useful for what ? One speaker proposed that there were two ways in which mathematics was useful. There was the traditional role of mathematics as a service subject to other disciplines, where its utility is evident. But there was also the idea that mathematics was useful to society, and that mathematics education had the
295
additional sociological benet of preparing students to be effective members of society, and of being able to contribute to the development, management, and efcacy of society. For mathematics to be useful to society, added another speaker, it is not a matter of teaching mathematics so as to be useful, but of teaching useful mathematics, which is a different matter. Mogens Niss added that when making a distinction between pure and applied mathematics, we should make a distinction between the epistemological level and the sociological level. One problem is that, even if there is no difference at the epistemological level, there is a distinction at the sociological level. And the problem is that at school (and also at university) this distinction is being perpetuated. In his view, the distinction is that you are an applied mathematician when the question comes from outside and you want an answer ; you are a pure mathematician if, when you are inspired by problems that come from outside, you seek to provide some sort of generality of solutions. APPLIED MATHEMATICS IN THE MATHEMATICS CURRICULUM. There was a general view that the applications of mathematics deserved a place within the school curriculum. Many have argued for this for a long time and had they been successful in bringing applicable mathematics into the school curriculum we would not be having this discussion at all and the intended ICMI Study into applied mathematics would not be necessary. Certainly we need a lobby group when we see that the distance between what we have and what we want is so great. This in itself would warrant an ICMI Afliated Study Group on modelling. The curriculum needs to be penetrated to the extent that the difference between pure and applied mathematics simply disappears. There are projects which seek to cross the borders between mathematics and other disciplines. Andr Revuz referred to an initiative at Paris VII university e where the teaching of a course was shared by mathematicians and physicists in the rst year of DEUG (Diplome d tudes universitaires g n rales). Each e e e week, in a one hour session the same question was presented by both teachers (who did not always agree). All were happy with this (and the students were delighted). To talk to people of different disciplines is to enrich both disciplines. If you ask the general public what it is in mathematics that is useful, you nd references to elementary mathematics, perhaps with sophisticated applications (elementary ratio, probability, statistics, etc.), or else undergraduate level mathematics. It is very rare to nd evidence of applications of upper secondary school mathematics. So a reason why applications do not penetrate school mathematics at that level is simply because there are none : very
296
GENERAL DISCUSSION
few people make extensive use of, say, trigonometric identities or quadratic equations. Of course, the greatest difculty in doing applied mathematics is the translation of the problem from the real context to a mathematical task (and this is true at all levels). The reason why applied mathematics is so difcult is because there are no givens and no correct answers. There is also the problem that validation of the solution of a problem lies outside the mathematical context (see the discussion on assessment later). The challenge for mathematics is that mathematics for all is hard to achieve. However there is a great deal of mathematics being taught outside mathematics lessons in other disciplines and if we wish to determine the extent of penetration of mathematics into other disciplines we need to look at what is being taught there. One consequence of the penetration of mathematics into other disciplines is that the mathematics there is being taught by practitioners of these other disciplines ; economists teach the maths needed for doing economics, physicists for physics, etc. If mathematicians are not willing to cross subject boundaries and engage with the teachers of these other subjects, what then happens to the teaching of mathematics ? If all the useful mathematics is no longer taught in mathematics classrooms by mathematics teachers, what happens to the mathematics left for the mathematics classroom ? The mathematician becomes a teacher of mathematics that leads nowhere. And the real danger then is that mathematics attains the status that Latin had in the 20th century part of our cultural heritage, but gradually abandoned as a useful tool for living in todays society. Trevor Fletcher, a retired inspector of schools in England and Wales, saw that the distinction between cultural mathematics and utilitarian mathematics was important but deplorable. Certainly at school level, mathematics should see itself as having the task of explaining something, rather than as contemplating itself. In fact, he claimed, the only justication for mathematics being compulsory as a school subject was that it had the power to explain. A nal point made was that motivation is an important factor in learning mathematics, and applications of mathematics can be important in encouraging curiosity. ASSESSMENT. The role played by assessment in the mathematics curriculum has important implications for the teaching of applications of mathematics. One speaker pointed out that teaching does not take place in an abstract context but in a reality which is the current school reality. Here assessment plays a signicant role and the reason why more applications of mathematics are
297
not taught can be explained simply by the fact of examinations at all levels. When examinations assess applications of mathematics, it is only sterilised and formalised presentations of the applications of mathematics that are examined. We can only penetrate school syllabuses if we take up the challenge : how can it be examined ? Against that, Djordje Kadijevi reminded the conference c that there were examples of experiments in assessment of applications of mathematics, at least as regards the use of computer assisted systems (CAS), which have been used in Austria for about 10 years. Mogens Niss was asked to say some more about applied mathematics and the problems of assessment. The problem for schools and universities, he said, was that they do not want the students to go all the way with tackling a problem. One problem is that validation of the solution requires expertise from outside the mathematics classroom. Another is that assessment carries important implications of costs (in time and resources). At his own university (Roskilde University Center), the students tackle applications of mathematics through topic-based work. Typically, though not exclusively, the mathematics application would be in a science subject and there would be joint supervision. Otherwise there would be a primary supervisor, but then the student is always free to go to others for advice, which is freely given. Certainly the overall costs in time and resources are expensive compared with traditional examining, but this is a cost they are willing to bear. This can be done because the university is new, and so there are no entrenched assumptions, and also because it is small. Perhaps really radical things can only be done on a small scale. Mogens Niss asked us to think of his situation as a laboratory, where practices are tried out which others can observe and then adapt the practices to their own institutions. It is not his view that the model should be copied as it stands.
STAKES IN MATHEMATICS EDUCATION FOR THE SOCIETIES OF TODAY AND TOMORROW Les enjeux de l ducation math matique e e pour les soci t s daujourdhui et de demain ee par Ubiratan DAMBROSIO
On examine dans cette contribution l mergence des math matiques comme une e e ` discipline dans les syst` mes scolaires a partir du XIXe si` cle. La pr occupation e e e ` des educateurs de nombreux pays visant a offrir un enseignement math matique e actualis dominait les r exions sur les objectifs, les contenus et les m thodes de e e e cet enseignement. ` Ce mouvement s tendit a tous les pays et devint ouvertement international. Cest e dans ce cadre que la revue LEnseignement Math matique fut cr ee a Gen` ve en 1899 e e ` e et que la Commission internationale de lenseignement math matique fut etablie par e ` le Congr` s international des math maticiens a Rome en 1908. Des soci t s nationales e e ee surgirent dans plusieurs pays. Apr` s la Seconde Guerre mondiale, les processus de d colonisation et la n cessit e e e e ` dinstruire les jeunes pour dautres genres de travail en les pr parant a une soci t de e ee consommateurs o` ils soient conscients de l tat du monde, tant du point de vue de u e lenvironnement que de la soci t , de l conomie et de la g opolitique, requi` rent ee e e e une nouvelle r exion sur une education math matique regroupant de nouvelles e e ` mati` res r sultant dune pens e interdisciplinaire et m me transdisciplinaire, suite a e e e e ` e leffondrement de barri` res culturelles et a l tablissement de nouvelles techniques e de communication et dinformation. Les objectifs, les contenus et les m thodes e de linstruction math matique subissent des transformations profondes, dont cette e contribution tente de tenir compte pour un enseignement math matique renouvel . e e
STAKES IN MATHEMATICS EDUCATION FOR THE SOCIETIES OF TODAY AND TOMORROW by Ubiratan DAMBROSIO
THE
SCENARIO IN THE TRANSITION FROM THE
19TH
TO THE
20TH
CENTURY
The transition from the 19th to the 20th century was marked by the effects of the three major revolutions of the Modern World : the Industrial Revolution, responsible for new models of production and labor relations ; the American Revolution, establishing a new model of governance and political leadership ; and the French Revolution, focusing on new social relations, advancing the modern concept of citizenship and generating new demands for bureaucracy and administration. In geopolitics, the 19th century established the great empires which emerged from the colonial order. At the same time, new intellectual and material instruments, developed on the basis of Newtonian science, were the bases for the establishment of the new sciences of sociology and economics. There was also the noticeable emergence of new directions in Christianity, to a great extent proposing new perceptions of man and society. History was seen as a determinant of the state of the world. The sciences, particularly mathematics, were consolidating the directions proposed since the 17th and 18th centuries. Rigor, precision, correctness were seen as major goals to be attained, both for the advancement of knowledge and as attributes of a valuable personality. Technology was an evidence of the righteousness of these pursuits. The Eiffel Tower, the Solvay Institute, the Nobel Prizes, are symbols of the achievements of modern technology. The world order was apparently tranquil. The results of the Raj revolt and of the Zulu War, the successes of the Creoles (Americans born of European descent) in establishing independent mirror-nations in the Americas, were all indicators of this tranquility.
STAKES IN MATHEMATICS EDUCATION
303
This World order implied the recognition of some universals. Mathematics was rmly established as a symbol of universal knowledge. And mathematicians from the entire world decided to assemble. An International Congress of Mathematicians was held in Zurich in 1897. Its membership reveals the dominating concept of universality : 197 members from 15 European countries plus 7 members from the USA. Another symbol of universality would be Esperanto, a universal language. Its structure, grammar, and vocabulary also reveal the prevailing concept of universality. Some mathematicians adhered to Esperanto. The second International Congress of Mathematicians was convened in Paris, in 1900. The 232 participants came from 26 countries. Non-Europeans 1 ) came from Algiers (1), Argentina (1), Brazil (1), Canada (1), Mexico (1), Peru (2), USA (8). The congress was marked by the conference of David Hilbert delivered in the section on History and Pedagogy, entitled simply Mathematical Problems, in which he formulated 23 problems which would mark the development of mathematics and, of course, Mathematics Education, throughout the 20th century. In the second paragraph of his paper, Hilbert says :
History teaches the continuity of the development of science. We know that every age has its own problems, which the following age either solves or casts aside as protless and replaces by new ones. [] the close of a great epoch not only invites us to look back into the past but also directs our thoughts to the unknown future. [Browder 1976, 1]
Almost one hundred years later, Stephen Smale would also refer to the interest of dealing with mathematical problems :
mathematics is more like art than other sciences. But there is one special difference, I nd, which is that mathematics tends to be correct. Mistakes in mathematics are rarely signicant for very long. But also mathematics tends to be more irrelevant. There is so much of mathematics that tends to go off in directions which are appreciated only by a very few, irrelevant even to the rest of mathematics. So I think there is a bigger danger in mathematics than there is in other sciences of tendencies to go off into irrelevancies, i.e., into things that are correct but not important. [Casacuberta & Castellet 1992, 8889]
Not only correctness, interest, relevance, appreciation, have always been present in reections about mathematics and mathematics education, but the utility of mathematics has also been emphasized. In the meeting of the British
1 ) In the next congress, held in Heidelberg in 1904, there was a larger number of participants (336), but only 18 came from outside Europe : 15 from the United States, 1 from Argentina, and 2 from Japan. In that same year, Japan would surprise the world by defeating the powerful Russian navy.
304
U. DAMBROSIO
Association, called to discuss the reform proposed by John Perry, his concept of usefulness is, in itself, a program [Perry 1901, 45]. Its eight points focus on mental pleasure, on logical styles of thinking, on preparing for further studies and professions. Although Perrys focus was received with criticism and hostility 2 ) by many mathematicians, his concept of usefulness has since then been recurrent in discussions about why teach mathematics. Perry is but an example of the movement towards a renewed mathematics education on the eve of the 20th century. Even at the risk of being accused of banalizing the discussions about Education, I claim that the discussions about curriculum focus on why, what and how to teach, or objectives, contents and methods [DAmbrosio 1983]. Much emphasis has been given to contents and methods. Objectives have essentially been taken for granted.
MATHEMATICS
EDUCATION REFORM
In the second part of the 19th century, the social order was showing signs of fragility. The call for emancipation by the popular classes and by women had obvious reections in education 3 ). The success of the United States in building up a new nationality showed the potential of schools for social change. The claim for social change was strong. The developments of a new production system, as a result of the Industrial Revolution, called for a new kind of worker. The home, so important when artisanal production was prevailing, did not respond to the needs of the future worker. The relation educationsociety was profoundly affected by two factors : the need for a new kind of worker ; the great progress of psychology, with the recognition of specic behavior and special needs of children. These two factors, the development of productive social capabilities and the modern discovery of the child are, according to Manacorda [1989], characteristic of the new school emerging in the second half of the 19th century. The new educational thinking called for updated mathematical contents, as well as for new methods of teaching, reecting ndings in learning. An active school was proposed : hands-on projects, learning by doing. Motivation should have priority.
2 ) hostility is the tone of the reaction of F. A. Forsyth, as recognized by himself [Perry 1901, 39]. 3 ) Quite interesting are the remarks of Mrs. W. N. Shaw on supporting mathematics teaching for girls : those powers in which the feminine mind is said to be peculiarly decient the powers of accurate observation and logical reasoning [Perry 1901, 50].
305
Mathematics Education was central for the new needs of labor. This is seen in Perrys and in several other proposals, where we recognize an appeal to psychology. In 1905 the Young couple wrote :
We have heard it asserted that it would be harmful to a young child to be plagued with Geometry. Geometry, as we have in view, is no plague. It requires no school-room, no blackboard, no special apparatus. It does not even require a trained teacher, nor does it demand the childs attention for long periods at a time. It is essentially a subject for home, and for an early age. [Young & Young 1905, vi]
And, resorting to paper folding, there follows good traditional Euclidean geometry. A few years later, Klein wrote :
The child cannot possibly understand if numbers are explained axiomatically as abstract things devoid of content, with which one can operate according to formal rules. [] While this goes without saying, one should mutatis mutandis take it to the heart, that in all instruction, even in the university, mathematics should be associated with everything that is seriously interesting to the pupil at that particular stage of his development and that can in any way be brought into relation with mathematics. [Klein 1908, 4]
A conservative posture, encouraged by the recent theoretical advances of classical mathematics and its applications, pointed to a formal and structured organization of curricula. Objectives, aimed at keeping the established social and world order, were taken for granted. Contents were almost entirely agreed upon, methods would differ widely. Particularly interesting is Kleins argument in favor of calculating machines [Klein 1908, 1722]. LEnseignement Math matique was founded in 1899 for contact, exchange e of information and comparison of educational systems. I will not comment on the rst series of the journal. This has been done in a very thoughtful and complete way by others. I will just add a few remarks, which may be taken as curiosities or may suggest reections of a different nature : 1) all the so-called general papers are written in French, except one in Esperanto, by M. Fr chet (1913), and a tiny few 4 ) in English, by C. Runge e (1912), D. E. Smith (1912), A. N. Whitehead (1913), and G. A. Miller (1915) ; 2) there are only two papers by Felix Klein, one in 1906 and another, joint with A. Gutzmer, in 1908, both about reforms in Germany ;
4 ) Of course English, German, Italian, and even Esperanto, do occur in other texts, in particular in reports of discussions.
306
U. DAMBROSIO
3) very few articles came from South America, though there is one by Nicolas Besio Moreno 5 ), from Argentina (1920) ; 4) in H. Fehrs report [1953] on the Commission internationale de lenseignement math matique, the USSR is not mentioned a single time, but there e is a reference to a paper on mathematics education in Russie (sic), by D. Sintsof 6 ). According to Manacorda [1989], many educational reforms in the Soviet Union, affecting particularly the objectives of education, and mathematics education in particular, were taking place. In the International Congress of Mathematicians, held in Rome in 1908, an International Commission on the Teaching of Mathematics was created by initiative of David Eugene Smith, from the USA, with the main objective of comparing methods and plans of studies of different countries. A retrospective of this commission was given by the outgoing chairman, H. Fehr, in 1952 [Fehr 1953]. In the same year, 1952, the General assembly of the International Mathematical Union (IMU), convened in Rome, established the International Commission on Mathematical Instruction (ICMI), as a commission of the IMU.
AFTEREFFECTS
OF
WORLD WAR II
The Second World War represents a landmark in recent world history. It paved the way for a new world and social order. The emergence of a new technology, called technoscience, strongly grounded in new areas of mathematics, asked for profound reforms of mathematics education. This was generally called Modern Mathematics Movement. New priorities for economic reconstruction and defense prompted the adoption of new ideas in Mathematics Education, anchored on advances in the theory of cognition. The research conducted by Jean Piaget was a strong support in the Western world, a role played by A. R. Luria and Lev Vygotsky in the Soviet Union. A major consequence of World War II was the establishment of the United Nations, in 1945. The charter was signed by 51 states. Soon it was enlarged by former colonies that achieved or were granted independence. The ideal of education for all was explicit in the Universal Declaration of Human Rights, issued by the United Nations in 1948. International cooperation was channeled through UNESCO, the Organization of American States, and similar organizations, as well as bilateral cooperation, establishing a new professional
5 6
) a renowned engineer who was well informed on educational matters. ) LEnseign. Math. 32 (1933), 8187.
307
international gure : the expert, or consultant, or coop rant. The rhetoric of e development implied the priority of mathematics over other disciplines in the curriculum. The Modern Mathematics Movement was seen as a valid option also for the nations belonging to the group called Third World 7 ). I will not engage in pro and con arguments about the Modern Mathematics Movement. Instead, I will address social and cultural issues, which were largely disregarded in the movement. The Cold War, immediately following the end of World War II, was marked by reconstruction of the economies, together with enormous spending in defense, which also provided funds for scientic research. Mathematics particularly beneted from generous military funding. At the same time, peace movements were active and the appeal of a new social and political order was intense. In the Third World, this appeal provided a fertile ground for civil wars and dictatorships, and a superb arena for the Cold War. The postwar period was also marked by intense demographic change. Granting independence was, in most cases, accompanied by very loose immigration procedures by most colonial metropoli. Emigration became a goal. The more developed nations were subjected to a large number of legal and illegal immigrants. This was added to the demand for civil rights, which was particularly strong in the USA. Social exclusion, intimately associated with cultural differences, emerged as a new focus of attention. Multicultural education, particularly language education, became a new issue. Indeed, children, and even adults, may acquire, in schooling, different kinds of cultural identity, thus perceiving different orders of understanding themselves and the world. The social implications of language education, recognized as a major factor responsible for exclusion, drew the attention of educators [Freire 1970 ; Bernstein 1971]. The reections focusing on language led scholars to examine the nature of languages and the cognitive implications of different languages. It was accepted that language was a cultural manifestation. Different cultures have different languages. The social implications of not respecting language differences were easily recognized. But few would dare similar thinking about mathematics. In the International Congress of Mathematicians held in Oslo in 1936, it was decided that the next congress would take place in the USA in 1940. As a consequence of the war, the congress was convened only in 1950, in Cambridge, Massachusetts. The International Mathematical Union (IMU) was
7 ) A general view of the movement, particularly focused on Brazil, can be seen in [B. S. DAmbrosio 1987].
308
U. DAMBROSIO
founded 8 ) in 1951 and its rst General Assembly took place in Rome in 1952. In this meeting, it was approved that an International Commission on Mathematical Instruction (ICMI), a reorganization of the pre-war International Commission on the Teaching of Mathematics, be formalized as a commission of IMU. After World War II, mathematics was rmly established as universal knowledge. The political atmosphere just after World War II conduced, with respect to mathematics, to what has been called the American Declaration of Universality, which proposed that mathematicians should convene independently of national allegiance [Lehto 1998, 74]. The universality implied objectives and goals for mathematics education dened irrespective of social and cultural parameters. In 1952 it was decided that ICMI should perform a study on the role of mathematics and mathematicians in the contemporary world. Duro Kurepa conducted the study, which was reported to the International Congress of Mathematicians in Amsterdam, 1954. The report gives special attention to several new directions of mathematical activity. Particularly interesting is the recognition of the importance of calculators :
Les math matiques se sont rapproch es de la physique et de la psychologie e e et on est en train dexaminer le m canisme du penser , du calculer et e dautres fonctions psychiques et intellectuelles. 9 ) [Kurepa 1955, 100101]
Recognizing the profound changes in the world, Kurepa gives a very comprehensive view of mathematics ; he sees its role, because of the imminence of the presence of robots, as essential for the mutual understanding between individuals and collectivities in order to have a global apprehension of the world. The role of mathematics in Weltanschauung is clearly stated as a moral duty of the mathematician. The questionnaire continued to produce interesting responses and reactions. The most intriguing, opening new perspectives for mathematics education, came from the renowned Japanese mathematician Yasuo Akizuki, a member of the Executive Committee of ICMI. In line with Kurepa, Akizuki proposed an emphasis on the reective side of mathematics, looking into the world as a whole. He made a strong point for introducing the history of science and mathematics in all levels of teaching. The most interesting point in his argument is the recognition that mathematics is a cultural product. He recognizes that Western mathematics is present in Asia, and says :
8 9
) For a full story of the I.M.U., see [Lehto 1998]. ) Mathematics gets closer to physics and to psychology and we are in the course of looking
into the mechanisms of thinking, of calculating and other psychic and intellectual functions.
309
Oriental philosophies and religions are of a very different kind from those of the West. I can therefore imagine that there might also exist different modes of thinking even in mathematics. Thus I think we should not limit ourselves to applying directly the methods which are currently considered in Europe and America to be the best, but should study mathematical instruction in Asia properly. Such a study might prove to be of interest and value for the West as well as for the East. [Akizuki 1959, 289]
Although anthropologists and psychologists had been showing interest in different kinds of mathematics, better saying, different ways of mathematizing in different cultures, Akizukis proposal did not attract the attention of the mathematical community until Claudia Zaslavsky published Africa Counts, in 1973 [Zaslavsky 1973].
NEW
DIRECTIONS OF MATHEMATICS EDUCATION REFORMS
The Royaumont Seminar, in 1959, convened by initiative of the European Economic Cooperation Administration, which later became the Organization for Economic Co-operation and Development (OECD), established the major guidelines for the movement known as Modern Mathematics. Policies for universal implementation of the recommendations called for regional cooperation. Marshall Stone, as the President of ICMI (19591962), marshaled cooperation in Mathematics Education in the Americas. The USA had been active in cooperation for development in the region, of vital importance in Cold War strategy, mainly through the Organization of the American States 10 ). In 1961 the First Inter-American Conference on Mathematics Education was held in Bogot , Colombia, largely nanced by the NSF. Prestigious a mathematicians from the USA and from Europe attended the conference, as well as representatives from almost every country of the Americas. As a result, the Inter-American Committee on Mathematics Education was founded, with Marshall Stone as its rst President. In 1966 a Second Inter-American Conference on Mathematics Education was held and since then, in different countries of the Americas, conferences were convened every four years [Barrantes & Ruz 1998]. The last one, the 10th IACME, took place in Maldonado, Uruguay, in November 1999.
10 ) Indeed, national research councils had been created in almost every country in the Americas, modeled on the National Science Foundation.
310
U. DAMBROSIO
The series of International Congresses on Mathematical Education (ICME) gave the possibility of IMU focusing more on contents regarding Mathematics Education. In 1971 the IMU created the Union Lectures, designed as an expository series of four to six lectures surveying current research themes. It was decided that the lectures would be published in LEnseignement Math e matique. This was a prompt response to acts that may be interpreted as weakening the relations of LEnseignement Math matique with ICMI. Hans e Freudenthal, while President of ICMI (19671970), sponsored the creation of Educational Studies in Mathematics, which obtained nancial support from UNESCO. The following President of ICMI, James Lighthill (19711974), created an ICMI Bulletin in 1972, as a means of spreading information about ICMI activities. It was considered to play a complementary role to that of LEnseignement Math matique, which remains the ofcial organ of ICMI. The e ICMI Bulletins appeal to mathematics educators all over the world increased steadily, as it attained regularity of publication and a large worldwide coverage of events and issues. Indeed, looking retrospectively on the issues covered by the ICMI Bulletin, as seen in issue no 47 (December 1999), we notice a very different character from LEnseignement Math matique. It is natural to e question the necessity and the reason for this dual approach. SCENARIOS 20TH 21ST
FOR THE TRANSITION FROM THE
TO THE
CENTURY
Before World War II, the objectives of mathematics education, aimed at keeping the established social and world order, were taken for granted. Contents were almost entirely agreed upon and methods would not differ substantially. Anchored in advances in the cognitive sciences and the new possibilities of calculation and information retrieval, the Modern Mathematics movement brought new contents and new methods of mathematics education. However, the objectives of mathematics education were unclear. And they remain so. The rhetoric of personal and social advances is not clearly focused, and exclusion seems to be the most noticeable effect of mathematics education. The question Why teach mathematics ? seems to be the crux 11 ). But together with this question come other questions about the nature of mathematics and how to handle mathematics teaching. Mathematics in the making ? Mathematics of everyday life ? Mathematics grounded in cultural traditions ? Mathematics as fun ? Good old classical mathematics ? Although not dichotomic, these strands do represent didactical options.
11
) I dealt with this question in ICME 3, Karlsruhe, 1976 [DAmbrosio 1979].
311
To face these recurrent questions, I see as fundamental a new, broader understanding of the socio-cultural history of mathematics and of its education. This calls for attention to non-conventional sources, rather than to purely academic ones. The Program Ethnomathematics [DAmbrosio 1992] provides the instruments to deal with these questions. The growing interest in ethnomathematics prompted the creation of an International Study Group in Ethnomathematics (ISGEm) in 1985. Since then, the publication, twice a year, of a Newsletter with international circulation, the realization of the First International Congress of Ethnomathematics, and the establishment of a web site 12 ) have givem worldwide visibility to ethnomathematics. An electronic journal, Mathematical Anthropology and Cultural Theory 13 ), has been recently created. As expected, criticism of ethnomathematics is mounting. Ethnomathematics was drawn into the Math Wars ! Recently The Chronicle of Higher Education started a very lively discussion on ethnomathematics. The discussions, which are going on, are an echo to the completely mistaken title of the article Good-bye, Pythagoras ! [Greene 2000]. Good old Pythagoras will ever be present to enable us to y and to communicate via the internet and many other important achievements of modern civilization. The male and female triangles of the Xingu culture will not do that. And, of course, there is no point in teaching a boy or girl in Chicago the way the Xingu culture classies triangles, except if shown as a piece of folklore, which indeed harms the proposal of ethnomathematics. Critics of ethnomathematics and, regrettably, even some supporters of ethnomathematics, are missing the point. Ethnomathematics does not propose to replace academic mathematics by folkloric mathematics, or Mickey Mouse mathematics, as it used to be called in the sixties. I am afraid the ethnomathematics proposal risks being seen and practiced ! in the same distorted way as the Modern Mathematics movement. Soon a modern Tom Lehrer will sell thousands of CDs ridiculing ethnomathematics, as Buffalo Bills Wild West Show did with Native American culture. The key issue is much deeper. It asks for a discussion of the major objectives of education and of schooling in the future. And, of course, how does mathematics t in this future. The difculty resides in a simple question : what do we know about the future ? Clearly, the way we see the future and the way we want the future guide our actions in the present.
12 13
) http://www.rpi.edu/~eglash/isgem.htm ) http://www.SBBay.org/MACT
312
U. DAMBROSIO
THE
FUTURE
I am not alone in dreaming of a future without hatred and bigotry, with peace among nations, peace in society, peace with the environment, individuals in peace with themselves. IMU is sponsoring World Mathematical Year 2000. The Assembly General of the United Nations has proclaimed 2001 to 2010 as the International Decade for a Culture of Peace and Non-violence for the Children of the World. The Assembly called on relevant United Nations bodies, non-governmental organizations (NGOs), religious bodies and groups, educational institutions, artists and the media to support the Decade actively for the benet of every child in the world. How do we respond to this call ? How does the IMU resolution and the UN resolution relate ? Shouldnt they be intrinsic to each other ? After all, mathematics is the dorsal spine of modern civilization. To give priority to peace may be a dream, maybe utopia, which justies my efforts as an educator and my joy as a grandfather, who hopes to survive to become a great-grandfather ! The future generations must live in a better world than the one which our and the previous generations before us have constructed. What can we offer to the future generations ? Not our model. But a critical view of this model and of the knowledge system in which it was built. Mathematics is recognized as central to this knowledge system. Hence, mathematics, and its history, are subjected to this critical view 14 ). It is important to understand some characteristics of the so-called echoboom generation (those born between 1977 and 1997). The boom generation, responsible for much of the expansion of the educational systems in the sixties and for the important events of 1968, is now reaching retirement age. The effects of this retirement play an important role in public nances, hence on politics. Much of my data refers to the USA but can easily be extrapolated to the developed world, where educational decisions set the standard. The echoboom generation is the largest ever, about 80 million young people, spanning from 3 to 23 years old, with an enormous purchasing power. Cinema and television currently the object of political disputes over who controls them are of lesser importance for the echo-boomers. Instead, they access the Internet and they feel control over it. Effectively, they have control, as hackers demonstrate. They benet from the technology gap between generations, which include parents, teachers, politicians, executives and decision makers in general.
14 ) This was a motivation for my proposal of editing a section for the Zentralblatt f ur Didaktik der Mathematik on Mathematics, peace and ethics. It appeared in Zbl. Didaktik Math. 30 (June 1998), Heft 3.
313
The economy, work and personal relations will be the most affected. They will be in control in a couple of decades. Most probably they will not be co-opted by the system, as it happened with the rebels of the sixties. The echo-boomers are creating a new culture 15 ). How will mathematics t in this new culture ? We need to understand how, in other epochs, mathematics was affected by changing scenarios. This justies my insistence on history of mathematics, not as understanding the development of a corpus of knowledge, but as a response to societal changes. Does society inuence 16 ) the development of mathematics ? Throughout history, in every culture, we recognize the efforts to develop instruments : i) to communicate ; ii) to cope with reality ; iii) to understand and to explain reality, providing the tools of critical thinking ; iv) to dene strategies for action. These goals can be identied in every ethnomathematics 17 ). Greek mathematics focused mainly on iii) and iv). The propdeutical character of mathematical education, which has been intrinsic to curriculum development in the 20th century, emphasized ii). In face of the new facilities of computation and information retrieval, there is no place for the propdeutical character of mathematical education. Numbers, gures, signs are communicative instruments, enriching the capability of discourse and conversation, of description. Critical familiarity with them, embedded in diversied cultural environments, is part of dealing with communicative instruments. Cultural environments mean calculators and computers if they are around, beads and threads if they are around, paper, pencil, chalk and blackboard if they are around. To create a school environment detached from the socio-cultural environment is justied only if projecting into the future, like the use of ction as a pedagogical tool. Discontinuities between school and out-of-school environments must be accompanied by a critical reection, not as a teaching device. Teaching goes on more and more in out-of-school environments. Both mathematics and ethnomathematics provide instruments to socialize the quantitative and qualitative ways of dealing with the surrounding reality. It is a responsibility of schools to prepare students to generate new realities, prospective or imaginary, that is, to be creative ; to be able to explain and understand reality with the capability of moving into the future equipped with
15 16
) See an interesting article in the journal The Futurist 34 (5) (Sept.Oct. 2000).
) Quite provocative is the paper by Loren R. Graham Do Mathematical Equations Display Social Attributes ? [Graham 2000].
17
) See [Urton 1997].
314
U. DAMBROSIO
strategies of action. This requires abstraction, conceptualization, in essence the domain of analytic instruments. Both mathematics and ethnomathematics provide such instruments, as history shows us. Technology is part of our world, and will be even more so in the future. From birth through death, we improve the capabilities of our body, the distribution of biological or sky time, our reachable and productive space, as well as our capability of communicating with living and dead individuals and cultures, through technology. When, why and how artifacts and instruments can be used, combined, improved and invented, that is, critical familiarity with material instruments that are part of modern civilization, already or potentially accessible, is an important objective of education. Both mathematics and ethnomathematics are intrinsic in such instruments, as history shows us. In the educational systems of the future I see mathematics, as well as ethnomathematics, inbuilt in the effort of critically providing communicative, analytic and material instruments 18 ). Production and labor will be present in every model of society of the future. The forms they will take surely will reveal the presence of high technology in everyday life. Jobs, as we understand them today, will most probably disappear 19 ). The social access of minorities is directly related to their acquiring communicative, analytical and material instruments, in other ways, to the implementation of the trivium LITERACYMATHERACYTECHNORACY.
IN
GUISE OF CONCLUSION
Is mathematics, as we today understand it in our curricula, prone to disappear ? I believe so. Curricular organization and assessment, the pillars of current school mathematics, reveal their fragility. The propdeutical character of mathematics, responsible for curricular organization, is challenged. Signs of this are the proliferation of remedial courses and of self-contained specic training courses, such as those provided by enterprise universities. World War II revealed the efciency of task-training, in line with Skinners proposals. Good achievement in mathematics, as assessed by standardized and similar varieties of tests, such as SIMS, TIMSS and others, seems to be unrelated to society
18
) Elsewhere [DAmbrosio 1999] I called these respectively literacy, matheracy and techno-
racy.
19 ) For a discussion of labor in the future, see [Reich 1992]. Harsh views of the future of employment, revealing the inadequacy of current educational systems, can be read in [Forrester 1999].
315
development. The downfall of the Soviet Union and the critical view of education in the Asian countries are signs of this. What will be the role of a mathematics teacher ? I believe mathematics teachers, understood as those who simply teach mathematics, will disappear. But there will be need for teachers who do mathematics. Even doing mathematics using only addition of three-digit numbers ! The teacher of the future will be a resource-companion of students in the search for new knowledge 20 ). The more mathematics a teacher knows, a better resource he/she will be ; the more curiosity about the new, a better companion he/she will be. The characteristics of the new teacher cannot be only the result of special training. Teachers of old generations will retire and echo-boomers will become the new teachers. I trust they will be able to help build a better world. How does this LEnseignement Math matique t in this scenario ? Surely, e there is always room for a good, respected, mathematical journal in this future. But does it fulll the needs of a new enseignement math matique ? Not in e the current format. Maybe, following the vocation of its founders who clearly understood what was going on in the transition from the 19th to the 20th century and were able to fulll the need of a journal with the characteristics of LEnseignement Math matique , a second rebirth might give rise to a new transdisciplinary e and transcultural journal of mathematics and mathematics education, opening its pages to historical, anthropological, and cultural issues.
REFERENCES AKIZUKI, Y. Proposal to I.C.M.I. LEnseign. Math. (2) 5 (1959), 288289. BARRANTES, H. y A. RUZ. La Historia del Comit Interamericano de Educacion I e Matem tica / The History of the Inter-American Committee on Mathematical a Education. Acad. Colombiana de Ciencias Exactas, Fsicas y Naturales, Santa F e de Bogot , 1998. a BERNSTEIN, B. Class, Codes and Control. I. Theoretical Studies towards a Sociology of Language. Routledge & Kegan Paul, London, 1971. BROWDER, F. E. (ed.) Mathematical Developments Arising from Hilbert Problems. Proceedings of Symposia in Pure Mathematics, Vol. 28. Amer. Math. Soc., Providence (R.I.), 1976. CASACUBERTA, C. and M. CASTELLET. (eds.) Mathematical Research Today and Tomorrow : Viewpoints of Seven Fields Medalists. Springer-Verlag, Berlin, 1992. [The passage quoted is from the Round-Table Discussion : pp. 87108.]
20
) I owe much to Eliot Wiggintons project Foxre for this approach.
316
U. DAMBROSIO
DAMBROSIO, B. S. The Dynamics and Consequences of the Modern Mathematics Reform Movement for Brazilian Mathematics Education. Ph. D. thesis. Indiana University, Bloomington, 1987. DAMBROSIO, U. Overall goals and objectives of mathematics education. In : New Trends in Mathematics Teaching IV, Chap. IX, 180198. UNESCO/ICMI, Paris, 1979. Successes and failures of mathematics curricula in the past two decades : A developing society in a holistic framework. In : Proceedings of the Fourth International Congress on Mathematical Education, ed. Marilyn Zweng et al., 362364. Birkh user, Boston, 1983. a Ethnomathematics : a research program on the history and philosophy of mathematics, with pedagogical implications. Notices of the Amer. Math. Soc., 39 (1992), 11831185. Literacy, Matheracy, and Technoracy : A trivium for today. Mathematical Thinking and Learning 1 (2) (1999), 131153. FEHR, H. [1953] Commission internationale de lenseignement math matique. LEnseign. e Math. 39 (19421950), 162168. FORRESTER, V. The Economic Horror. Routledge, New York, 1999. FREIRE, P. Pedagogy of the Oppressed. Seabury, New York, 1970. GRAHAM, L. R. Do mathematical equations display social attributes ? Math. Intelligencer 22 (2000), 3136. GREENE, E. Good-Bye Pythagoras ? (Ethnomathematics embraces non-European methods of math ; critics fear a decline in rigor.) The Chronicle of Higher Education (6 Oct. 2000), A16 [http://chronicle.com/free/v47/i06/06a01601.htm]. KLEIN, F. [1908] Elementary Mathematics from an Advanced Standpoint. Arithmetic, Algebra, Analysis. Translated by E. R. Hedrick and C. A. Noble (from the 3 rd German edition, 1924). Dover Publications, New York. ` e KUREPA, G. Le r le des math matiques et du math maticien a l poque contemporaine. o e e Rapport g n ral. LEnseign. Math. (2) 1 (1955), 93111. e e LEHTO, O. Mathematics Without Borders. A History of the International Mathematical Union. Springer-Verlag, New York, 1998. ` MANACORDA, M. A. Storia delleducazione dallantichita a oggi. Edizioni Rai, Torino, 1989. PERRY, J. (ed.) Discussion on the Teaching of Mathematics. (A British Association Meeting at Glasgow, 1901.) Macmillan and Co., London, 1901. REICH, R. B. The Work of Nations : Preparing Ourselves for 21st Century Capitalism. Vintage Books, New York, 1992. URTON, G. The Social Life of Numbers. A Quechua Ontology of Numbers and Philosophy of Arithmetic, with the collaboration of Primitivo Nina Llanos. University of Texas Press, Austin, 1997. YOUNG, G. C. and W. H. YOUNG. [1905] A Beginners Book of Geometry. Chelsea Publishing Company, New York, 1970 (the same, except for corrections, as the original edition of 1905). ZASLAVSKY, C. [1973] Africa Counts. Number and Pattern in African Cultures. Third edition. Lawrence Hill Books, Chicago, 1999.
SCIENTIFIC SOLIDARITY TODAY AND TOMORROW La solidarit scientique aujourdhui et demain e par Jeremy KILPATRICK
Les soci t s modernes int` grent les math matiques dans les programmes scolaires ee e e ` pour des raisons a la fois pratiques et intellectuelles. Historiquement, les math matiques e e ont et enseign es au niveau de l cole primaire pour fournir des outils de r solution des e e e probl` mes pratiques, tandis quelles sont entr es dans le syst` me educatif secondaire e e e ` et sup rieur a travers les humanit s (liberal arts). Depuis le milieu du 20e si` cle, des e e e e enjeux pratiques et intellectuels pour lapprentissage des math matiques ont et vis s e e ` dans la plupart des pays et a tous les niveaux de la scolarit . A la n de ce si` cle, e e les enjeux pratiques ont pris une telle importance par rapport aux enjeux intellectuels que ces derniers risquent d tre eclips s. e e A limage du 20e si` cle, qui a vu une expansion mondiale de l ducation e e ` secondaire universelle, il est probable quau 21e si` cle on assistera a une expansion e similaire du secteur universitaire avec une demande importante de math matiques e scolaires. Nos soci t s ont accru la quantit de math matiques quelles attendent en ee e e vue de comp tences num riques g n rales et, de ce fait, retardent le moment o` e e e e u beaucoup d tudiants commencent leurs etudes sp cialis es conduisant aux math e e e e matiques avanc es. La transition entre l ducation secondaire et l ducation sup rieure e e e e en math matiques va requ rir une attention particuli` re. e e e La deuxi` me moiti du 20e si` cle t moigne de l mergence et de la disparition aussi e e e e e bien de projets curriculaires concernant les math matiques que de nombreux centres e nationaux d ducation math matique. La CIEM devrait etudier comment de tels centres e e peuvent etre renforc s, et favoriser non seulement le d veloppement et la recherche sur e e les programmes mais aussi la formation professionnelle continue dont les enseignants ` ont besoin. Un d majeur pour le si` cle a venir est daugmenter la solidarit entre e e e ` ` chercheurs et enseignants de math matiques a tous les niveaux (de l cole primaire a e e luniversit ). e
SCIENTIFIC SOLIDARITY TODAY AND TOMORROW by Jeremy KILPATRICK
I should like to discuss questions of the stakes in mathematics education for the societies of today and tomorrow : For whom is mathematics taught, and why ? What political tools (in their broadest sense) should we mathematics educators develop for reecting on how to teach ? These are not minor questions, and with limited space to discuss them, I clearly can address only some aspects of each one. I have decided to concentrate on the question of why societies include mathematics in the school curriculum and for whom it is taught. Regarding tools for reection, I have only a few possibly provocative comments to make at the end.
1. WHY
TEACH MATHEMATICS ?
Over the years, a variety of justications have been offered for teaching mathematics 1 ). These justications can be classied in a variety of ways. For example, some are concerned with why every modern society has made mathematics a compulsory part of schooling, whereas others concern the reasons individual students might have for studying mathematics. In this discussion, I address only the former, the reasons that societies provide mathematics instruction to their members. I make a further simplication : there are essentially two broad categories of reasons for people to learn mathematics the practical and the intellectual. Practical justications for teaching mathematics range from a societys need to have a numerate citizenry, one that can cope with the quantitative
1
) For reviews and more elaborated discussions, see [DAmbrosio 1979; Niss 1996; Stanic
1984].
SCIENTIFIC SOLIDARITY TODAY AND TOMORROW
319
demands of everyday life in the home, marketplace, and workplace, to its need for people who can build its buildings and bridges, develop its seeds and medicines, and design its airplanes and computers. Societies invest in school mathematics because they want to improve their technological and socio-economic standing, as Niss points out, either as such or in competition with other societies/countries [1996, 13]. They have political, economic, and even military reasons for wanting some of their members to know and be able to use advanced mathematics in solving a variety of practical problems. They also want as many people as possible to be numerate at some basic level. Intellectual justications for teaching mathematics range from a societys need for an educated citizenry that understands and appreciates the role mathematics has played in building that society and developing its culture to its need for people who can extend mathematics into new realms. Deductive reasoning, proof, and axiomatic structure, for example, are features of mathematics that ought to be understood and appreciated by educated people in a modern society. Mathematics has historically been taught at least in part because people have believed that its study develops habits and attitudes that are important for their intellectual development. Even today, a cogent argument can still be made that all students, and not just those who go on to become mathematicians and create new mathematical ideas, can prot intellectually and aesthetically from studying the mathematics developed in their society and elsewhere. These two types of justication are connected with the way in which mathematics has been institutionalized in schools across the centuries. All societies, as they have developed methods for symbolizing ideas, have also developed their own mathematical systems, which have then had to be taught to the next generation. Todays school mathematics curricula can be seen as resulting in large part from the collision of two tectonic plates. The rst plate developed within primary education. Societies have traditionally established some form of primary education as a means of giving children a rudimentary education in reading, writing, and elementary arithmetic. Since the mid-nineteenth or early twentieth century, free, universal, and compulsory primary education has been available in the industrialized countries of the world, and the second half of the twentieth century saw a tremendous expansion of primary education in developing countries. Mathematics is virtually universal in the curriculum of primary education, with the traditional role of mathematics teaching in primary schools being to prepare children for their future societal roles. In principle, they were learning mathematics as a tool for solving practical problems.
320
J. KILPATRICK
In contrast, the second tectonic plate developed within Western secondary and tertiary education. When the seven liberal arts were rst formulated around 100 BC, mathematics cutting across the four liberal arts of arithmetic, geometry, music, and astronomy had already acquired a prestigious place in the curriculum of the academy. For Plato, mathematics was the test for the best minds. Secondary and higher education should aim not at the accumulation of knowledge but rather at the development of the intellect. Because mathematics was included among the liberal arts, it remained alive in the monastic and cathedral schools of medieval Europe and in the universities that were their successors even when little original research was being done in mathematics and the level of mathematical teaching was not very high. As new universities were founded during the nineteenth century around the world and the older universities expanded and developed in Europe, mathematics in secondary education was directed toward university preparation. It retained its liberal arts character. In a simplied view, therefore, the tectonic plates of justication for the traditional school mathematics curriculum looked roughly as follows :
Primary
Secondary
Primary school mathematics aimed at teaching the practical side of mathematics, and, for those who continued their education, secondary school mathematics aimed at the intellectual side. This rough picture of tectonic plates characterized many curricula at the beginning of the twentieth century, and in some countries it is still a reasonable portrayal. As the gure might suggest, the middle grades marking the transition between primary and secondary education became one of the most contentious parts of the curriculum as societies struggled to organize school mathematics at those grades to allow a smooth transition from one set of goals to another. It is no accident that the middle grades were characterized in the mid-twentieth century as the doldrums of school mathematics, nor that many of the curriculum reform efforts that became known as the new math began in the middle grades.
321
As part of those reform efforts, although it had begun earlier, the tectonic plates began to shift in many societies. Mathematics in the primary grades acquired some of the intellectual character of secondary mathematics, and, as more secondary students and teachers embraced practical reasons for the study of mathematics, it became a more practical subject in the secondary grades. The line separating the practical and intellectual justications for studying mathematics began to cut across grades rather than to separate levels of education, as below :
Primary
Secondary
The shift in the line of demarcation became even greater in the last three decades of the twentieth century as technology made the study of applications of mathematics and mathematical modeling not only more feasible but also more desirable for those students who would go on to use mathematics in their careers to solve practical problems. Despite its many oversimplications, this analysis allows us to see the following : 1. For centuries, societies have had both practical and intellectual justications for teaching mathematics, and these remain intertwined in todays school mathematics. 2. The twentieth century witnessed a broadening of the mathematics curriculum at all levels, which allowed practical and intellectual aims to cross more grades. 3. In the last three decades, more arguments have been put forward for practicality in school mathematics, and students now have more opportunities to use in realistic situations the mathematics they are learning. What is at stake for the societies of tomorrow are the intellectual justications for school mathematics : Can they be maintained in the face of what appears to be an overwhelming push for practicality ? In 1904, Poincar was bemoaning the extraordinary fact that so many people nd e
322
J. KILPATRICK
mathematical denitions and proofs obscure and emphasizing the importance to mathematicians of making their subject comprehensible to engineers. In todays societies, politicians, parents, students, and many mathematics teachers clamor for mathematics that students can use to solve practical problems, whether they become engineers or not. The old claims for the intellectual and aesthetic value of learning mathematics are seldom heard today except among mathematicians and some mathematics teachers who continue to argue for the disciplinary value of the subject. If one sees pure mathematics as the consequence of intellectual justications and applied mathematics as the consequence of practical justications, it appears as though the pure side of school mathematics is gradually being eclipsed by the applied side. Will the future see a total eclipse or a reappearance from somewhere of justications for developing the intellect ? Perhaps we will again one day hear the argument that there is nothing so practical as a good mathematical abstraction.
2. FOR
WHOM IS MATHEMATICS TAUGHT ?
The twentieth century saw the expansion of education at all levels in virtually every country of the world. In many respects, it might be called the century of the secondary school. Before the Second World War, secondary education, especially at the upper level, and to an even greater extent higher education, was only open to limited numbers of the corresponding age-group, and usually only to boys [UNESCO 2000, 14]. By the end of the century in many countries of the world, the opportunity for some form of secondary education had in effect been extended to all students. A recent report from UNESCO summarized the situation for developing countries :
[In the past half-century,] school enrolment has doubled or even tripled in the developing countries, according to the level of education. By the middle of the 1950s, [the] groundswell had reached secondary education, where structures had been unied (creation of comprehensive middle-schools), arriving in the 1970s at post-secondary and higher education, the generalization of which i.e. extension to over half of a generation is a reality today in many countries. [UNESCO 2000, 15]
The previous gures showing justications for the curriculum, with school mathematics sitting within a rectangle, clearly do not portray who takes mathematics. First, one should remember that even as access to secondary education has improved dramatically, there are still many countries of the world in which the number of students completing secondary school is well
323
less than half of their age cohort. If we were to draw a picture of the fraction of the cohort in school at each grade, and therefore available to be taught school mathematics, we would have, for each country, a gure roughly approaching a trapezoid or triangle and not a rectangle. Second, every society has to deal with built-in tensions that arise from the disparate goals of, rst, educating all students to be numerate and, second, educating some students to study mathematics at advanced levels. Modern societies want essentially everyone to learn basic mathematics as part of the preparation they need for adult life. Not everyone, however, needs or wants to study mathematics at the advanced levels that society must demand of some if that society is to develop and maintain its scientic and technical prowess. Society needs what has been termed a mainline to provide numerate citizens, and it needs a pipeline to provide specialists in mathematics. Of course, this picture is highly oversimplied, since societies have many ways of setting up mainlines and pipelines for mathematics, and they have often developed different pipelines for students headed for different careers that require a knowledge of advanced mathematics.
MAINLINE
PIPELINE
Regardless of how society has structured its educational system, however, at some point in each pupils education, a decision needs to be made as to whether, and if so when, the pupil will leave the school mathematics mainline to enter a school mathematics pipeline. When that decision is made, who makes it, and what the consequences will be for the pupil are questions that every society has had to face. One reason mathematics has come in for so much scrutiny over the past century has been its role in most societies as the lter or gatekeeper that determines who enters which pipelines since they lead not only to advanced study of mathematics but also to differentially desirable careers and social advantages. In some cases, the decision point to enter a school mathematics pipeline comes at the transition from elementary to secondary education, and the kind
324
J. KILPATRICK
of secondary school the pupil enters will determine whether he or she is in a pipeline. In other cases, the decision is made within a secondary school as the student is encouraged or not to pursue studies leading to advanced mathematics. And in still other cases, no decision is made until the student enters tertiary education and chooses or is chosen for a career demanding further study of mathematics. The general tendency during the twentieth century, as I see it, was to delay the decision point and sometimes to allow the decision to be reversed later if necessary. The decision has often been delayed more than in the past because of changes in what it means to be numerate. The sheer amount of mathematics that students need to know for numeracy today has been redened upward in many countries. Furthermore, the criteria used in making the decision have changed somewhat in many societies. Traditionally, performance on an examination in mathematics has been the main way to decide who stays in the mainline and who enters a pipeline. The quality of the examination has not always been high, the criterion for passing the examination has usually been set quite arbitrarily, and at times the connection between the examination and the school curriculum has been questionable. But schools have almost always found themselves needing to devote substantial time to preparing students for such examinations. In some educational systems during the twentieth century, the judgment of the mathematics teacher began to play a more important role, as did, in some cases, the desires of the pupil and his or her parents. As we consider challenges for the societies of the future, we can expect to see access to the study of mathematics continuing to expand around the world. A reasonably safe prediction is that if the twentieth century was the century of secondary education, the coming century will be the century of tertiary education. Note that I say tertiary education and not university education or university-level education. In many societies, the university is being supplemented by a variety of postsecondary institutions that lead to careers for which university studies would be neither necessary nor suitable. Universities as we know them today are only a couple of centuries old, although of course their roots go back several millennia. Change seems inevitable for universities over the next century, but the direction and magnitude of that change are far from clear. A number of commentators [Brown & Duguid 1996 ; Duderstadt 1997 ; Noam 1995 ; Reid 1998] view the twentyrst century university as likely to retain many if not all of its traditional functions but also to change in sometimes drastic ways under the pressure of electronic technology. In particular, Eli Noam, an economist, sees the economic foundation of universities collapsing because face-to-face teaching
325
is becoming too expensive. In contrast, John Seely Brown and Paul Duguid see a hybrid university emerging, in which institutional arrangements are changed and the university devolves into components a degree-granting body, the academic staff, the campus facilities, and the students that are no longer tied tightly together. Brown and Duguid envision a middle way between the natural centralizing tendencies of the monolithic university and the vision advanced by so many futurists of a completely dispersed distance education mediated by technology. In all of these scenarios, many more people enter universities and related institutions as tertiary education evolves into a more democratic and open global knowledge industry [Duderstadt 1997]. The challenge to school mathematics, regardless of which scenario comes about, will be to prepare increasing numbers of students to enter tertiary education equipped to continue learning not only mathematics but also the other subjects they will study. That challenge is nontrivial in view of the experience of the past century, which suggests that a rapid expansion of enrollments in advanced courses in school mathematics is almost inevitably accompanied by a perceived decline in the content and rigor of those courses 2 ). Creating a smooth transition from secondary to tertiary education so that societys intellectual and practical goals for learning mathematics are kept in some sort of balance is part of the challenge. LEnseignement Math matique was founded in the spirit of scientic e solidarity that brought together mathematicians and teachers of mathematics in a mutual effort to improve the teaching and learning of mathematics. From the beginning, the editors saw both the intellectual and the practical demands that would be placed on the subject of mathematics. As they said in the editorial in the inaugural issue :
Lavenir de la civilisation d pend en grande partie de la direction desprit que e recevront les jeunes g n rations en mati` re scientique ; et dans cette education e e e scientique l l ment math matique occupe une place pr pond rante. Soit au ee e e e ` point de vue de la science pure, soit a celui des applications, le XXe si` cle, e qui va souvrir, manifestera des exigences auxquelles personne ne doit ni ne peut se d rober. 3 ) [Fehr & Laisant 1899, 5] e
2 ) For further discussion of the challenge of ensuring mathematics for all, as well as other challenges facing mathematics educators in the United States and Canada, see [Kilpatrick & Silver 2000]. 3 ) The future of civilization depends mainly on the direction that the younger generations will receive in understanding scientic subject matter, and in this scientic education the mathematical element occupies a dominant place. Whether from the point of view of pure science or from that of applications, the approaching twentieth century will impose requirements that must and cannot be concealed from anyone.
326
J. KILPATRICK
Their words are as pertinent for the twenty-rst century as they were for the twentieth.
3. TOOLS
FOR REFLECTING ON HOW TO TEACH
Scientic solidarity has been demonstrated in a different way in the invention of curriculum development projects in mathematics. Although they may seem to have been around forever, curriculum development projects are actually a phenomenon of the last half of the twentieth century. Before that time, schools had a mathematics curriculum, of course, but it had not been developed through a project. All through the twentieth century, groups were organized to look critically at school mathematics, study various curriculum problems, and make recommendations. The International Commission on Mathematical Teaching was a pioneer in that process, stimulating a series of national reports early in the century. Most of the groups working on curriculum, however, conned their activities to the production of reports and recommendations that were seldom accompanied by teaching materials or efforts to try out the recommendations [Howson, Keitel & Kilpatrick 1981, 68]. The idea that the curriculum could be studied and purposefully developed, rather than just being allowed to evolve, led ultimately to the rst project in 1951 with the establishment of the University of Illinois Committee on School Mathematics at the dawn of the new math era. The curriculum development project, which was modeled after military and public health projects aimed at a specic result a weapon, a vaccine, a therapy , brought together mathematicians and teachers of school mathematics in a new kind of partnership and solidarity. One of the great lessons taught by the new math era, however, was that curriculum development and solidarity within a project are not enough. Every teacher is involved in curriculum development at some level, and if ideas and materials developed by projects are to take root in classrooms, teachers need to be educated and supported in becoming more competent, autonomous developers and users of curricula [Howson, Keitel & Kilpatrick 1981, 260]. That is why I see, as an essential item on the agenda for improving the practice of mathematics teaching in the coming century, the creation of new forms of continuous professional development for teachers of mathematics. It is far from clear what these forms might be or how they might be organized institutionally. The past half-century saw the establishment in many countries of centers and institutes where curriculum development
327
could take place. In some cases, these centers were able to undertake, in addition to curriculum development, programs of research in mathematics education, but most of them with the striking exception of the Instituts de recherche sur lenseignement des math matiques (IREMs) in France e did not have the resources, or even the mission, to do very much in the way of professional development. Furthermore, many of these centers have had a precarious existence. Governments have changed, bringing into ofce administrators who were not convinced that such centers warranted their continued support. Foundations have turned their funding elsewhere, guring that they had launched an effort that needed to continue on its own. Ofcials of universities that housed a center have wanted it brought under their control or wanted it terminated as an unnecessary drain on their resources. I cannot tell you how many letters I have been asked to write and have written over the past several decades to administrators threatening drastic cuts of one sort or another in which I testied to the good work a center had done and the need for its continued existence. These centers have sometimes been able to nd new sources of funding or have been able to reinvent themselves, but often they have continued on in a very diminished form, and sometimes they have essentially disappeared from view. Last January, I was present in Gothenburg for the inauguration of the Swedish National Center for Mathematics Education (NCM). I was delighted to see this enterprise underway with careful planning, talented people, and what appeared to be a rm commitment from the national government for sustained support but I also could not help worrying, in view of the history of similar ventures elsewhere, what the center might look like a decade from now. Any enterprise that is set up to address the continuous professional development of mathematics teachers, in my view, cannot avoid also addressing matters of curriculum development and of research. That is an ambitious program for anyone to undertake, and it is not clear that the type of center structure available in the past will continue to work. In particular, pressures to move such centers out from under the umbrella of the university are likely to increase in many countries. Somehow, ways need to be found to insulate such enterprises from the winds of politics both academic and civic. A useful activity for the International Commission on Mathematical Instruction (ICMI) or some other organization to take on would be to survey systematically the history and fate of institutes and centers for mathematics education around the world over the past three decades or so to see what might be learned from the various ways in which work has been organized, funding secured, and programs conducted. Such a survey might suggest how new
328
J. KILPATRICK
institutional structures might be built to house the sort of programs I envision for integrating professional development with curriculum development and research. In 1992, ICMI began a study entitled What is research in mathematics education, and what are its results ? The purpose, as expressed in the letter of invitation to conduct the study, was to review the state of the eld and to begin a dialogue with other scientic communities [Sierpinska & Kilpatrick 1998, x]. Those communities included, specically, the community of mathematicians. The ICMI Executive Committee felt that mathematicians did not know enough about what was happening in research in mathematics education and were questioning whether it had yielded any results at all worth considering. Anna Sierpinska and I edited the report of the study, which appeared in 1998 and was entitled Mathematics Education as a Research Domain. Because we wanted to convey the point that the study had not yielded a single, denitive answer to the questions it addressed, we made what turned out to be a tactical error and added the subtitle A Search for Identity. This resulted in reviewers making wisecracks about adolescents seeking identity and asking what other eld would be questioning whether or not it is a research discipline. Mathematicians, of course, would never engage in such questioning. Perhaps the questioning comes from the history of the eld, and in particular, from the way mathematics education has developed internationally, as illustrated in the ICMI. One can also ask, what other eld would have the ofcers of its premier international organization appointed by a group outside the eld ? The ICMI is a commission of the International Mathematical Union (IMU) and is therefore subject to IMUs oversight. Could it be that the insecurity and apparent disarray of the eld, despite its growth and accomplishments through the twentieth century, might stem in part from the way it has been treated by mathematicians ? In 1984, in an address at the Fifth International Congress on Mathematical Education, I made the following modest proposal :
Perhaps the eld [of mathematics education] has reached a point in its development where it needs to set up a permanent executive a secretariat that would facilitate communication among mathematics educators around the world. [] The International Commission might even wish to sponsor some sort of individual membership organization, possibly with a newsletter, so that interested persons might maintain contact with one another in the four years between congresses. [Kilpatrick 1985, 20]
I felt then, as I feel now, that the ICMI needed more autonomy as well as more involvement by its constituency. Since that time, I have served on
329
the ICMI Executive Committee, and I have a better understanding of the reasons, not only historical but also economic, that have kept the ICMI from becoming or initiating a membership organization despite the growth of the eld. Nonetheless, I believe that the ICMI needs to become an independent group of some sort. I dont want to be misunderstood. I am certainly not suggesting that mathematics educators should sever all ties with mathematicians. Both groups have much to learn from each other. I have recently been very encouraged, despite the so-called math wars (cf. [Schubring 2003, 63]), to encounter mathematicians who are beginning to understand better than in the past and truly to appreciate the complexity and difculty of the issues that mathematics educators are wrestling with in their research and their practice. For too long, mathematics education has been seen as a eld open to anyone knowledgeable about mathematics in which having an interest and perhaps some experience in teaching is sufcient for entry. That attitude is, I think, beginning to change. I used to think that it would be helpful to mathematics education if it could be seen as included among what are called the mathematical sciences. I have changed my mind. The tactic is a little like the one used during the so-called back-to-basics movement of the 1970s in North America : dene problem solving as a basic skill and then embrace the slogan back to basics. Such rhetorical tricks never work. Mathematics education is not a branch of mathematics, nor does it belong among the arts and sciences. It is a separate eld with very different traditions, foundations, problems, methods, and results. It is much more contingent on history and culture than mathematics could ever be, and that is part of the reason for what outsiders perceive as a eld in disarray. A major challenge for the twenty-rst century is for mathematicians and mathematics educators to modify their mutual relationship. Moreover, both need to modify their relationship with teachers. In a recent book, Ellen Lagemann [2000] documents the widening gap during the twentieth century between US education researchers and school-teachers. The condescension researchers in mathematics education have shown teachers of school mathematics has sometimes mirrored the condescension they have received themselves from mathematicians. All three groups need to learn to work more productively together with greater understanding of their differences and greater appreciation for the strengths the others are bringing to the enterprise. The relationships need to shift from outmoded paternalism to true fraternalism. Only then can we have scientic solidarity in learning and teaching mathematics as well as with mathematics itself.
330
J. KILPATRICK
REFERENCES BROWN, J. S. and P. DUGUID. Universities in the digital age. Change : The Journal of the American Academy of Higher Education 28 (1996), 1019. [http://www.parc.xerox.com/ops/members/brown/papers/university.html] DAMBROSIO, U. Overall goals and objectives of mathematical education. In : H.-G. Steiner and B. Christiansen (eds.), New Trends in Mathematics Teaching. Vol. 4, 180198. UNESCO, Paris, 1979. DUDERSTADT, J. J. The future of the university in an age of knowledge. Journal of Asynchronous Learning Networks 1 (2) (1997), 7888. [On-line] [http://www.aln.org/publications/jaln/v1n2/v1n2 duderstadt.asp] FEHR, H. et C. LAISANT. LEnseignement math matique. LEnseign. Math. 1 (1899), e 15. HOWSON, G., C. KEITEL and J. KILPATRICK. Curriculum Development in Mathematics. Cambridge University Press, 1981. KILPATRICK, J. Reection and recursion. Educational Studies in Mathematics 16 (1985), 126. KILPATRICK, J. and E. A. SILVER. Unnished business : Challenges for mathematics educators in the next decades. In : M. J. Burke and F. R. Curcio (eds.), Learning Mathematics for a New Century, 223235. National Council of Teachers of Mathematics, Reston (VA), 2000. LAGEMANN, E. C. An Elusive Science : The Troubling History of Education Research. University of Chicago Press, 2000. NISS, M. Goals of mathematics teaching. In : A. J. Bishop, K. Clements, C. Keitel, J. Kilpatrick and C. Laborde (eds.), International Handbook of Mathematics Education, 1147. Kluwer, Dordrecht, 1996. NOAM, E. M. Electronics and the dim future of the university. Science 270 (1995), 247249. [http://www.columbia.edu/dlc/wp/citi/citinoam14.html] POINCARE, H. Les d nitions g n rales en math matiques. LEnseign. Math. 6 (1904), e e e e 257283. REID, W. A. Erasmus, Gates, and the end of curriculum. Journal of Curriculum Studies 30 (1998), 499501. [http://faculty.ed.uiuc.edu/westbury/JCS/VOL30/reid.html] SCHUBRING, G. [2003] LEnseignement Math matique and the rst International Come mission (IMUK) : the emergence of international communication and cooperation. This volume, 4765. SIERPINSKA, A. and J. KILPATRICK. Mathematics Education as a Research Domain : A Search for Identity. Kluwer, Dordrecht, 1998. STANIC, G. M. A. Why teach mathematics ? A historical study of the justication question. (Doctoral dissertation, University of Wisconsin-Madison, 1983.) Dissertation Abstracts International 44 (1984), 2347A. UNESCO. 50 Years for Education (2000). [On-line] [http://www.unesco.org/education/educprog/50y/brochure/index.htm]
REACTION The growing importance and challenges of mathematics education by Hyman BASS
Both Jeremy Kilpatrick and Ubiratan DAmbrosio offer us a grand historical perspective on mathematics education, reaching back into the 19 th century. Perhaps appropriately, they dwell more on yesterday and today than on tomorrow. DAmbrosio emphasizes the social and cultural context of mathematics. While both authors treat the question Why teach mathematics ?, they offer different kinds of analyses. Discussing the transition from the twentieth to the twenty-rst centuries, DAmbrosio writes :
Anchored in advances in the cognitive sciences and the new possibilities of calculation and information retrieval, the Modern Mathematics movement brought new contents and new methods of mathematics education. However, the objectives of mathematics education were unclear. And they remain so. [] The question Why teach mathematics ? seems to be the crux. But together with this question come other questions about the nature of mathematics and how to handle mathematics teaching. Mathematics in the making ? Mathematics of everyday life ? Mathematics grounded in cultural traditions ? Mathematics as fun ? Good old classical mathematics ? [DAmbrosio 2003, 310]
The Ethnomathematics Program, of which DAmbrosio is a founder, addresses these questions in a socio-cultural and historical context.
Throughout history, in every culture, we recognize the efforts to develop instruments : i) to communicate ; ii) to cope with reality ; iii) to understand and to explain reality, providing the tools of critical thinking ; iv) to dene strategies for action. [DAmbrosio 2003, 313]
He sees the role of mathematics education in different historical periods as emphasizing different subsets of these four characteristics.
332
H. BASS
In the educational systems of the future I see mathematics, as well as ethnomathematics, inbuilt in the effort of critically providing communicative, analytic and material instruments. [DAmbrosio 2003, 314] The teacher of the future will be a resource-companion of students in the search for new knowledge. The more mathematics a teacher knows, a better resource he/she will be ; the more curiosity about the new, a better companion he/she will be. The characteristics of the new teacher cannot be only the result of special training. Teachers of old generations will retire and echo-boomers will become the new teachers. I trust they will be able to help build a better world. [DAmbrosio 2003, 315]
Kilpatricks perspective is more internal to the culture of mathematics as a discipline, and to the infrastructure of the mathematics research community and of the mathematics educational enterprise. Broadly characterizing the genres of missions of mathematics education as either practical or intellectual, he suggests that, in the early twentieth century, practical aims dominated primary education, while intellectual aims prevailed at the secondary level. As the century progressed, both practical and intellectual aims emerged as important at both levels. Gradually, stimulated in part by technology, the practical assumed increasing dominance over the intellectual, a development that he notes with some lament. He further observes that, in the twentieth to twenty-rst century transition, tertiary (i.e. post-secondary, but not necessarily university) level education would become more widespread and important. Another motif concerns the bifurcation in mathematics education between general students (mainline), and those needing advanced mathematical study for mathematically intensive professions (pipeline). Kilpatrick notes that the bifurcation point, where one course of study or the other is decided, is evolving to a later point in time. Moreover, the exam-based criterion for these two pathways is being progressively broadened. Kilpatrick offers some important observations and views concerning curriculum development, professional development of teachers, and education research. He calls attention to a remarkable historical fact :
The idea that the curriculum could be studied and purposefully developed, rather than just being allowed to evolve, led ultimately to the rst project in 1951 with the establishment of the University of Illinois Committee on School Mathematics at the dawn of the new math era. The curriculum development project, which was modeled after military and public health projects aimed at a specic result a weapon, a vaccine, a therapy , brought together mathematicians and teachers of school mathematics in a new kind of partnership and solidarity. [Kilpatrick 2003, 326]
333
Most mathematics education reform interventions have been founded on the development of new curricular materials. However, their implementation has often proved disappointing, an outcome judged by many to have been a consequence of inadequate concern for the capacity-building that would have been needed to make these curricula usable and effective in school. Kilpatrick refers to [Howson et al. 1981] :
Every teacher is involved in curriculum development at some level, and if ideas and materials developed by projects are to take root in classrooms, teachers need to be educated and supported in becoming more competent, autonomous developers and users of curricula. [Kilpatrick 2003, 326]
He draws a conclusion that has become now almost universally acknowledged :

That is why I see, as an essential item on the agenda for improving the practice of mathematics teaching in the coming century, the creation of new forms of continuous professional development for teachers of mathematics. [Kilpatrick 2003, 326]
And he further argues that

Any enterprise that is set up to address the continuous professional development of mathematics teachers, in my view, cannot avoid also addressing matters of curriculum development and of research. [Kilpatrick 2003, 327]
and that
ways need to be found to insulate such enterprises from the winds of politics, both academic and civic. [Kilpatrick 2003, 327]
As co-chair of an ICMI Study on Mathematics Education Research, Kilpatrick has reected at length on the character and status of this eld. With regard to its relation to mathematics, he insists that
Mathematics education is not a branch of mathematics [] It is a separate eld with very different traditions, foundations, problems, methods, and results. It is much more contingent on history and culture than mathematics could ever be, and that is part of the reason for what outsiders perceive as a eld in disarray. [Kilpatrick 2003, 329]
For reasons such as these he has advocated a fundamental revision of what are seen to be the anachronistic and patriarchal organization of the governance and of ICMI within the International Mathematical Union (IMU), and its attendant lack of autonomy. Kilpatrick notes for example that the ofcers and EC of ICMI are elected by the IMU, without voting participation by ICMI. This is an issue of continuing concern and debate, to which I am now a direct witness. It seems likely that progress will be evolutionary, not revolutionary, and this is perhaps to be desired.
334
H. BASS
In a concluding comment, Kilpatrick notes that mathematicians and mathematics education researchers are joined by a third professional community of practice, school mathematics teachers, and that the whole triangle of mutual relationships is in need of improvement.
The condescension researchers in mathematics education have shown teachers of school mathematics has sometimes mirrored the condescension they have received themselves from mathematicians. All three groups need to learn to work more productively together with greater understanding of their differences and greater appreciation for the strengths the others are bringing to the enterprise. [Kilpatrick 2003, 329]
SOME
AFTERTHOUGHTS
Both Ubi DAmbrosio and Jeremy Kilpatrick have provided us with invaluable historical canvases of the sweep of educational issues and the institutional forms of addressing them over the past century, and more. I am most familiar with the setting of Kilpatricks presentation, and I am in broad agreement with his analyses and recommendations. In these brief remarks, let me bring into sharper focus a few points that I feel are now particularly pressing. First of all, I cannot overemphasize the importance of continuing professional development of teachers. Developing frameworks and curriculum materials is crucial and challenging work, but it is far easier than building human capacity in a profession of human improvement like teaching. Any stable educational improvement at systemic scale depends in large measure on this. This requires growth in our knowledge base (research), and implementation takes time and sustained support from public institutions. Other countries, such as Japan, provide examples of a professional culture of teaching that supports continuous professional development. In every aspect of educational improvement, it must (tacitly if not explicitly) be understood that every recommendation must pay a tax to the need for concomitant professional development. This need is now widely acknowledged and appreciated. It is the subject of a recently launched ICMI Study, co-chaired by Deborah Ball and Ruhama Even. This applies in particular to the matters discussed below. Finally, let me open a bit wider the discussion of Why study mathematics ? As DAmbrosio emphasizes, the answers to this change with history and culture, and they are now conditioned also by the presence of powerful technology. Our unquestioned premise of the central place of mathematics in the school curriculum (often 12 years of instruction) is now under widespread
335
assault (in Europe, in Japan, and in some quarters in almost every country). This is producing policies that can have profound impact on the cultural, intellectual, and material standing of mathematics, and its professional communities, in the educational enterprise. The traditional rationale for mathematics study has been (to give it a ner grained classication than Kilpatricks practical and intellectual) a mixture of pragmatic, economic, social, intellectual, and cultural reasons. Pragmatic reasons include the need to learn the basic skills of arithmetic and of measurement, and the rudimentary geometric concepts and gures. Economic arguments are based on the quantitative literacy demanded by the rapidly evolving technological workplace, and the desire to remain competitive in the world economy. Mathematics can be justied for social reasons, because it provides the resources for responsible citizenship in a modern industrial democracy. The intellectual justication is that mathematics is the enabling discipline for all of science, and that it offers fundamental tools of analysis, quantitative expression, and disciplined reasoning. Finally, its cultural warrant is that mathematics exposes students to some of the most subtle and sublime achievements of the human spirit. Yet people can still question whether these arguments sufce to justify so many years of mathematical study. Who today will be doing signicant arithmetic without a calculator ? How many people will ever have occasion as adults to solve a quadratic equation ? Even engineers have no need to prove that their mathematical methods work. Why do our cultural arguments for the study of mathematics merit such different responses than those advocating the study of ancient languages and history ? Most of my colleagues, mathematicians and teachers, have strong personal convictions about the importance of a solid mathematics education for all citizens, but these are increasingly seen to be expressions of belief systems rather than persuasive, empirically-based, arguments for public policy. So the rst step in convincing policy makers is to convince ourselves. The answers must be framed in new ways, and grounded in new conceptions of what a contemporary mathematics curriculum, and learning goals for all students, must look like. For example, one argument is that mathematics study cultivates analytical reasoning skills that are a powerful general intellectual resource. To this one might rst observe that the teaching of mathematical reasoning is an endangered species in much of current mathematical instruction, with widely observed disabling effects. Secondly, the implicit argument for knowledgetransfer is vividly challenged within the very domain of concern to us here
336
H. BASS
mathematics education. Highly competent mathematicians have been frequently guilty of making strong general claims and judgments about mathematics education that are empirical in nature, yet based on limited and often anecdotal evidence. They exhibit a kind of cavalier undisciplined reasoning in mathematics education that they would never countenance in mathematics itself. Questions about the place and nature of mathematics education comprise a web of complex and subtle relationships, but they are crucial to the evolution of mathematics education, and so they demand our serious attention and reection. Here is not the place to enter into the important issues that these questions bring to the fore. Some aspects of this set of questions are being engaged by a project on quantitative literacy, led in part by Lynn Steen, who is chief author of a discussion document, Mathematics and Democracy : The Case for Quantitative Literacy, published in 2001 by the Mathematical Association of America. There are few contemporary questions about mathematics education more important than those concerned with the kind, scope, and nature of the mathematics needed by all responsible citizens in a modern technological democracy, and how such education can be realized.
REFERENCES DAMBROSIO, U. [2003] Stakes in mathematics education for the societies of today and tomorrow. This volume, 301316. HOWSON, G., C. KEITEL and J. KILPATRICK. Curriculum Development in Mathematics. Cambridge University Press, 1981. KILPATRICK, J. [2003] Scientic solidarity today and tomorrow. This volume, 317330.
Imprim par SRO-KUNDIG SA Gen` ve e e

100 Ans Enseignement Mathematiques

Загружено:

Сведения о документе

Оригинальное название

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

100 Ans Enseignement Mathematiques

Загружено:

Авторское право:

Доступные форматы

One Hundred Years of LEnseignement Math matique e

Moments of Mathematics Education in the Twentieth Century

Proceedings of the EMICMI Symposium Geneva, 2022 October 2000

c 2003 LEnseignement Math matique, Gen` ve e e

LENSEIGNEMENT MATHEMATIQUE : BIRTH AND STAKES

(reported by CHRIS WEEKS) . . . . . . . . . . . .

95 113 133 155 160

(reported by MARTA MENGHINI) . . . . . . . . .

179 191 211 224

(reported by MARTA MENGHINI) . . . . . . . . .

APPLICATIONS OF MATHEMATICS : MATHEMATICS AS A SERVICE SUBJECT

251 271 285 293

(reported by CHRIS WEEKS) . . . . . . . . . . . .

PERSPECTIVES FOR MATHEMATICS EDUCATION

301 317 331

Charles-Ange LAISANT (18411920)

Geneva in that same year !

) See in this connection LEnseign. Math. 21 (1920), 306.

Henri FEHR (18701954)

LENSEIGNEMENT MATHEMATIQUE : BIRTH AND STAKES

MATHEMATICAL INSTRUCTION IN AN INTERNATIONAL PERSPECTIVE

MATHEMATICAL INSTRUCTION IN AN INTERNATIONAL PERSPECTIVE

) with R. Fueter and M. Grossmann

on the mathematics teaching ;

FORMAT OF THE JOURNAL

MATHEMATICAL INSTRUCTION IN AN INTERNATIONAL PERSPECTIVE

MATHEMATICAL INSTRUCTION IN AN INTERNATIONAL PERSPECTIVE

MATHEMATICAL INSTRUCTION IN AN INTERNATIONAL PERSPECTIVE

THAT APPEARED IN THE JOURNAL, CLASSIFIED BY THEMES, AND THEIR DISTRIBUTION

) I refer to the modern names of the countries.

MATHEMATICAL INSTRUCTION IN AN INTERNATIONAL PERSPECTIVE

INQUIRY ON THE METHOD OF WORKING OF MATHEMATICIANS

MATHEMATICAL INSTRUCTION IN AN INTERNATIONAL PERSPECTIVE

MATHEMATICAL INSTRUCTION IN AN INTERNATIONAL PERSPECTIVE

MATHEMATICAL INSTRUCTION IN AN INTERNATIONAL PERSPECTIVE

1899, 1900, 1901

THE TOWN OF PUBLICATION THE FIRST TIME THE JOURNAL IS QUOTED)

MATHEMATICAL INSTRUCTION IN AN INTERNATIONAL PERSPECTIVE

NOTES ABOUT THE INQUIRY ON THE METHOD OF WORKING OF MATHEMATICIANS

MATHEMATICAL INSTRUCTION IN AN INTERNATIONAL PERSPECTIVE

MATHEMATICAL INSTRUCTION IN AN INTERNATIONAL PERSPECTIVE

Cours de mathmatiques accessoires

Mathmatiques spciales et Physique

THE FIRST INTERNATIONAL COMMISSION

THE FIRST INTERNATIONAL COMMISSION

) the concept of function, graphical representation of very simple functions.

THE FIRST INTERNATIONAL COMMISSION

Canada Cape Colony Chile China Egypt

Japan Mexico Peru Serbia Turkey

THE FIRST INTERNATIONAL COMMISSION

THE FIRST INTERNATIONAL COMMISSION

related exclusively to higher education. The titles of the reports 8 ) were :

THE FIRST INTERNATIONAL COMMISSION

THE FIRST INTERNATIONAL COMMISSION

THE FIRST INTERNATIONAL COMMISSION

JOURNALS OF MATHEMATICS EDUCATION, 19002000 by Gila HANNA

) using the description mathematics

JOURNALS OF MATHEMATICS EDUCATION, 19002000

JOURNALS OF MATHEMATICS EDUCATION, 19002000

JOURNALS OF MATHEMATICS EDUCATION, 19002000

THE KARLSRUHE ZENTRUM FUR DIDAKTIK

JOURNALS OF MATHEMATICS EDUCATION, 19002000

CLOSER LOOK AT THREE PROMINENT JOURNALS